2026-10-26 16:40:56 +00:00
16 changed files with 12716 additions and 396 deletions
@@ -2,9 +2,9 @@
 title: "Ai4Bio Schema Check"
 task: ""
 lineage_type: import
-upstream_source: https://github.com/inoue0426/awesome-computational-biology/blob/7a064bf0/.github/workflows/ai4bio-schema-check.yml
+upstream_source: https://github.com/inoue0426/awesome-computational-biology/blob/4ed7cfef/.github/workflows/ai4bio-schema-check.yml
-upstream_sha: 7a064bf0
+upstream_sha: 4ed7cfef
-imported_at: 2026-08-08
+imported_at: 2026-10-26
 prompt_class: unknown
 upstream_changes: accepted
 author: upstream
@@ -50,6 +50,7 @@ jobs:
      - name: Build enriched artifacts
        run: uv run --with pyyaml python scripts/build_resources_v2.py
      - name: Verify enriched artifacts are committed
        if: github.event_name == 'pull_request'
        run: |
          if git diff --quiet; then
            echo "AI4Bio artifacts are in sync."
@@ -2,9 +2,9 @@
 title: "Awesome Computational Biology [![Awesome](https://awesome.re/badge.svg)](https://awesome.re)"
 task: ""
 lineage_type: import
-upstream_source: https://github.com/inoue0426/awesome-computational-biology/blob/c6f07d90/README.md
+upstream_source: https://github.com/inoue0426/awesome-computational-biology/blob/4ed7cfef/README.md
-upstream_sha: c6f07d90
+upstream_sha: 4ed7cfef
-imported_at: 2026-08-31
+imported_at: 2026-10-26
 prompt_class: catalogue
 upstream_changes: accepted
 author: upstream
@@ -32,6 +32,22 @@ A curated collection of databases, software, and papers related to computational
 Browse and search the resources via the [GitHub Pages UI](https://inoue0426.github.io/awesome-computational-biology/).
 For treatment-response work, the [Dataset Explorer](https://inoue0426.github.io/awesome-computational-biology/dataset-explorer.html) compares curated cell-line, patient, and PDX datasets by year, sample type, perturbation metadata, matched pre/post availability, SMILES coverage, and clinical outcomes.
 The [Method Explorer](https://inoue0426.github.io/awesome-computational-biology/method-explorer.html) compares drug-response and perturbation methods by task, molecular/context representation, unseen-drug support, dose/time conditioning, and patient transfer.
 The [Foundation Model Explorer](https://inoue0426.github.io/awesome-computational-biology/foundation-explorer.html) compares foundation models by fine-grained modality, parameter count, pretraining scale, species, zero-shot support, weights/code availability, perturbation support, and spatial support.
 The [Agent Explorer](https://inoue0426.github.io/awesome-computational-biology/agent-explorer.html) compares agentic AI systems by scientific domain, single- vs multi-agent architecture, tool/code execution, literature and web retrieval, omics and wet-lab support, autonomy, and human-in-the-loop design. The Agent Explorer table is sortable and supports domain, architecture, omics, code-execution, and year filters.
 All explorer tables support sortable columns and focused filters for quick comparison, including a Recent (≥2025) toggle.
 The Pages home screen also summarizes total resources, profiled datasets, methods, foundation models, agents, and the most common foundation-model modalities.
 Chemical and genetic perturbation datasets can be filtered separately in the Dataset Explorer. Genetic screens can also be filtered by CRISPRi, CRISPRa, knockout, enhancer-targeting, combinatorial, or mixed perturbation modes, including mixed-resource collections.
 The explorer also includes the eight perturbation datasets used in the Bison unseen-compound benchmark.
 - Search matches `name`, `description`, `tasks`, `modalities`, and `tags`.
 - The **Task**, **Modality**, and **Type** filters map directly to `tasks`, `modalities`, and `type` in `docs/data/resources.json`.
 - Clicking badges on cards applies the corresponding filter.
@@ -236,44 +252,82 @@ Browse and search the resources via the [GitHub Pages UI](https://inoue0426.gith
 ## Benchmarks & Datasets
- [1000 Genomes Project](https://www.internationalgenome.org/) — Reference panel of human genetic variation from 2,504 individuals across 26 populations.
+### Drug Response & Perturbation
- [BACE](https://www.kaggle.com/datasets/gokturkkoch/bace) — Binary classification and regression dataset for β-secretase 1 (BACE-1) inhibitor binding affinity.
+
 - [BEAT AML](https://biodev.github.io/BeatAML2/) — Functional ex vivo drug sensitivity measurements paired with genomics for acute myeloid leukemia.
 - [Cancer Therapeutics Response Portal (CTRP)](https://portals.broadinstitute.org/ctrp/) — Drug sensitivity profiles across ~900 cancer cell lines for >400 compounds.
 - [Chem-PerturBridge](https://github.com/theislab/Chem-PerturBridge) — Harmonized compendium and processing pipelines for small-molecule perturbation transcriptomics across heterogeneous assays and gene panels.
 - [GSE191127 Breast Cancer Pre/Post Chemotherapy](https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE191127) — Bulk RNA-seq and genomics from matched pre- and post-neoadjuvant chemotherapy breast tumors with treatment outcomes.
 - [Genomics of Drug Sensitivity in Cancer (GDSC)](https://www.cancerrxgene.org/) — Drug sensitivity for ~1000 human cancer cell lines and hundreds of compounds.
 - [JUMP Cell Painting Datasets](https://github.com/jump-cellpainting/datasets) — Consortium-scale cell imaging perturbation datasets (chemical and genetic) for phenotypic profiling and drug discovery research.
 - [LINCS L1000](https://lincsproject.org/LINCS/tools/workflows/find-the-best-place-to-obtain-the-lincs-l1000-data) — Gene expression profiles (978 landmark genes) for >20,000 chemical and genetic perturbations across cell lines.
 - [LINCS L1000 Phase 1](https://github.com/theislab/Chem-PerturBridge) — Phase-1 L1000 benchmark view used by Bison with 692,787 profiles, 9,233 molecules, 70 contexts, and 978 landmark genes.
 - [LINCS L1000 Phase 2](https://github.com/theislab/Chem-PerturBridge) — Phase-2 L1000 benchmark view used by Bison with 333,263 profiles, 1,760 molecules, 30 contexts, and 978 landmark genes.
 - [MIX-Seq](https://www.nature.com/articles/s41467-020-17440-w) — Multiplexed single-cell transcriptional profiling of chemical and genetic perturbation responses across pools of cancer cell lines.
 - [NCI60](https://dtp.cancer.gov/discovery_development/nci-60/) — Drug sensitivity benchmark across 60 diverse human cancer cell lines.
 - [Novartis Perturbation Dataset](https://github.com/theislab/Chem-PerturBridge) — Large chemical perturbation transcriptomic screen used in Chem-PerturBridge; the Bison benchmark view contains 46,748 profiles and 3,770 molecules.
 - [OP3](https://github.com/theislab/Chem-PerturBridge) — Chemical perturbation dataset used by Chem-PerturBridge and the Bison unseen-compound benchmark; the Bison benchmark view contains 1,813 profiles, 138 molecules, and 4 cellular contexts.
 - [PDX-Atlas](https://muralportal.pdxatlas.org/) — Portal integrating clinical, genomic, transcriptomic, and drug-response data from patient-derived xenograft models.
 - [PharmGKB](https://www.pharmgkb.org/) — Curated pharmacogenomics dataset linking genetic variants to drug response phenotypes across thousands of drugs.
 - [PRISM](https://depmap.org/portal/prism/) — Cancer drug sensitivity profiling of >4,500 drugs across >900 cancer cell lines using pooled-cell-line barcoding.
 - [sci-Plex](https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE139944) — Single-cell chemical perturbation screen of ~650,000 transcriptomes across three cancer cell lines, 188 compounds, and four doses.
 - [scPerturb](https://github.com/sanderlab/scPerturb) — Curated and continuously updated single-cell perturbation data resource spanning CRISPR and drug perturbation studies.
 - [Tahoe-100M](https://huggingface.co/datasets/tahoebio/Tahoe-100M) — Giga-scale single-cell perturbation atlas with >100 million profiles from 50 cancer cell lines exposed to ~1,100 small molecules.
 - [VCPI-0001](https://github.com/theislab/Chem-PerturBridge) — Bulk chemical perturbation dataset used in Chem-PerturBridge; the Bison benchmark view contains 27,517 profiles and 2,272 molecules in one cellular context.
 - [VCPI-0002](https://github.com/theislab/Chem-PerturBridge) — Bulk chemical perturbation dataset used in Chem-PerturBridge; the Bison benchmark view contains 18,139 profiles and 1,488 molecules in one cellular context.
 ### Genetic Perturbation & Perturb-seq
 - [Dixit et al. 2016 Perturb-seq](https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE90063) — Pooled CRISPR knockout with single-cell RNA-seq across K562 and dendritic-cell screens, including stimulated conditions and combinatorial perturbations.
 - [Adamson et al. 2016 Perturb-seq](https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE90546) — CRISPRi Perturb-seq in K562 cells for systematic dissection of the unfolded protein response.
 - [Datlinger et al. 2017 CROP-seq](https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE92872) — Pooled CRISPR knockout screen with single-cell transcriptomic readout in Jurkat cells under T-cell receptor stimulation.
 - [Norman et al. 2019 CRISPRa Perturb-seq](https://figshare.com/articles/dataset/Norman_et_al_2019_Perturb-seq/27766323) — Large-scale CRISPR activation Perturb-seq in K562 cells spanning single-gene and combinatorial perturbations.
 - [Gasperini et al. 2019 CRISPRi enhancer screen](https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE120861) — High-MOI single-cell CRISPRi screen in K562 cells targeting thousands of candidate enhancers to map enhancer-gene regulation.
 - [Replogle et al. 2022 K562 Genome-wide Perturb-seq](https://plus.figshare.com/articles/dataset/_Mapping_information-rich_genotype-phenotype_landscapes_with_genome-scale_Perturb-seq_Replogle_et_al_2022_processed_Perturb-seq_datasets/20029387) — Genome-scale CRISPRi Perturb-seq targeting nearly all expressed genes in K562 cells.
 - [Replogle et al. 2022 K562 Essential Perturb-seq](https://plus.figshare.com/articles/dataset/_Mapping_information-rich_genotype-phenotype_landscapes_with_genome-scale_Perturb-seq_Replogle_et_al_2022_processed_Perturb-seq_datasets/20029387) — CRISPRi Perturb-seq focused on essential genes in K562 cells.
 - [Replogle et al. 2022 RPE1 Essential Perturb-seq](https://plus.figshare.com/articles/dataset/_Mapping_information-rich_genotype-phenotype_landscapes_with_genome-scale_Perturb-seq_Replogle_et_al_2022_processed_Perturb-seq_datasets/20029387) — CRISPRi Perturb-seq focused on essential genes in RPE1 cells.
 - [Frangieh et al. 2021 Perturb-CITE-seq](https://singlecell.broadinstitute.org/single_cell/study/SCP1064/multi-modal-pooled-perturb-cite-seq-screens-in-patient-models-define-novel-mechanisms-of-cancer-immune-evasion) — Multimodal CRISPR knockout screen with RNA and surface-protein readouts in patient-derived melanoma models under immune-related conditions.
 ### Single-Cell & Spatial
 - [HEST Xenium virtual spatial transcriptomics](https://huggingface.co/datasets/ratschlab/HEST_Xenium_virtual_spatial_transcriptomics) — DeepSpot-M predicted transcriptome-wide ST for 59 HEST-1k 10x Xenium samples (~13.3M cells) (gated). Paper: [DeepSpot-M](https://www.medrxiv.org/content/10.64898/2026.06.19.26356060v1).
 - [scIB (Single-cell Integration Benchmarks)](https://github.com/theislab/scib) — Comprehensive benchmarking framework for single-cell data integration methods.
 - [Tabula Muris](https://tabula-muris.ds.czbiohub.org/) — Comprehensive single-cell atlas of 20 mouse organs and tissues, enabling cross-tissue and cross-species comparisons.
 - [Tabula Sapiens](https://tabula-sapiens-portal.ds.czbiohub.org/) — Comprehensive human single-cell atlas of ~500K cells from 24 organs and tissues across multiple donors.
 - [TCGA virtual spatial transcriptomics atlas](https://huggingface.co/datasets/ratschlab/TCGA_virtual_spatial_transcriptomics_atlas) — DeepSpot-M predicted transcriptome-wide ST for TCGA H&E (FF + FFPE; 28,664 slides / 32 cancer types; gated). Paper: [DeepSpot-M](https://www.medrxiv.org/content/10.64898/2026.06.19.26356060v1).
 ### Molecular, Protein & Drug Discovery
 - [BACE](https://www.kaggle.com/datasets/gokturkkoch/bace) — Binary classification and regression dataset for β-secretase 1 (BACE-1) inhibitor binding affinity.
 - [Bento](https://github.com/LigandPro/Bento) — Protein-ligand docking benchmark covering rigid, flexible, de novo, blind, induced-fit, and covalent docking tasks.
 - [BindingDB Curated Sets](https://www.bindingdb.org/rwd/bind/chemsearch/marvin/SDFdownload.jsp?all_download=yes) — Curated binding affinity datasets for protein–ligand interaction benchmarking.
 - [Cancer Therapeutics Response Portal (CTRP)](https://portals.broadinstitute.org/ctrp/) — Drug sensitivity profiles across ~900 cancer cell lines for >400 compounds.
 - [ClinTox](https://tdcommons.ai/single_pred_tasks/tox/#clintox) — Clinical toxicity dataset contrasting FDA-approved drugs with those that failed clinical trials due to toxicity.
 - [CPTAC (Clinical Proteomic Tumor Analysis Consortium)](https://proteomics.cancer.gov/programs/cptac) — Multi-omic proteogenomic datasets for multiple cancer types linking proteomics with genomics.
 - [CrossDocked2020](https://arxiv.org/abs/2001.01037) — Large-scale dataset for structure-based virtual screening.
 - [DUD-E (Directory of Useful Decoys, Enhanced)](http://dude.docking.org/) — Structure-based virtual screening benchmark with active ligands and challenging decoy sets across diverse protein targets.
 - [FLIP (Fitness Landscape Inference for Proteins)](https://github.com/J-SNACKKB/FLIP) — Benchmark collection of protein fitness landscape datasets for evaluating protein ML models.
 - [Genomics of Drug Sensitivity in Cancer (GDSC)](https://www.cancerrxgene.org/) — Drug sensitivity for ~1000 human cancer cell lines and hundreds of compounds.
 - [GuacaMol](https://github.com/BenevolentAI/guacamol) — Benchmark suite for generative molecular design models.
 - [JUMP Cell Painting Datasets](https://github.com/jump-cellpainting/datasets) — Consortium-scale cell imaging perturbation datasets (chemical and genetic) for phenotypic profiling and drug discovery research.
 - [LINCS L1000](https://lincsproject.org/LINCS/tools/workflows/find-the-best-place-to-obtain-the-lincs-l1000-data) — Gene expression profiles (978 landmark genes) for >20,000 chemical and genetic perturbations across cell lines.
 - [MoleculeNet](http://moleculenet.ai/) — Benchmark datasets for molecular machine learning.
 - [MOSES](https://github.com/molecularsets/moses) — Benchmarking platform for molecular generation models.
 - [NCI60](https://dtp.cancer.gov/discovery_development/nci-60/) — Drug sensitivity benchmark across 60 diverse human cancer cell lines.
 - [OGB (Open Graph Benchmark)](https://ogb.stanford.edu/) — Large-scale graph ML benchmark suite including biological datasets such as ogbl-ppa (protein-protein associations) and ogbg-molhiv.
 - [OpenBioLink](https://github.com/OpenBioLink/OpenBioLink) — Benchmark datasets for biological knowledge graph completion.
 - [PharmGKB](https://www.pharmgkb.org/) — Curated pharmacogenomics dataset linking genetic variants to drug response phenotypes across thousands of drugs.
 - [PK-DB](https://pk-db.com/) — Open database of experimental pharmacokinetics (PK) and ADME data from clinical and preclinical studies.
 - [PRISM](https://depmap.org/portal/prism/) — Cancer drug sensitivity profiling of >4,500 drugs across >900 cancer cell lines using pooled-cell-line barcoding.
 - [ProteinGym](https://github.com/OATML-Markslab/ProteinGym) — Large-scale benchmark of deep mutational scanning assays for evaluating protein fitness landscape models.
 - [QM9](https://figshare.com/collections/Quantum_chemistry_structures_and_properties_of_134_kilo_molecules/978904) — Quantum chemistry properties for 134K stable small organic molecules computed at DFT level.
 - [scIB (Single-cell Integration Benchmarks)](https://github.com/theislab/scib) — Comprehensive benchmarking framework for single-cell data integration methods.
 - [scPerturb](https://github.com/sanderlab/scPerturb) — Curated and continuously updated single-cell perturbation data resource spanning CRISPR and drug perturbation studies.
 - [SIDER (Side Effect Resource)](http://sideeffects.embl.de/) — Database of 1,430 approved drugs with their recorded adverse drug reactions across 27 system-organ classes.
 - [Tabula Muris](https://tabula-muris.ds.czbiohub.org/) — Comprehensive single-cell atlas of 20 mouse organs and tissues, enabling cross-tissue and cross-species comparisons.
 - [Tabula Sapiens](https://tabula-sapiens-portal.ds.czbiohub.org/) — Comprehensive human single-cell atlas of ~500K cells from 24 organs and tissues across multiple donors.
 - [TAPE (Tasks Assessing Protein Embeddings)](https://github.com/songlab-cal/tape) — Benchmark suite of five biologically meaningful semi-supervised learning tasks for evaluating protein representations.
 - [The Cancer Genome Atlas (TCGA)](https://www.cancer.gov/about-nci/organization/ccg/research/structural-genomics/tcga) — Comprehensive multi-omics (genomics, transcriptomics, proteomics, methylation) dataset for 33 cancer types across ~11,000 patients.
 - [TCGA virtual spatial transcriptomics atlas](https://huggingface.co/datasets/ratschlab/TCGA_virtual_spatial_transcriptomics_atlas) — DeepSpot-M predicted transcriptome-wide ST for TCGA H&E (FF + FFPE; 28,664 slides / 32 cancer types; gated). Paper: [DeepSpot-M](https://www.medrxiv.org/content/10.64898/2026.06.19.26356060v1).
 - [HEST Xenium virtual spatial transcriptomics](https://huggingface.co/datasets/ratschlab/HEST_Xenium_virtual_spatial_transcriptomics) — DeepSpot-M predicted transcriptome-wide ST for 59 HEST-1k 10x Xenium samples (~13.3M cells) (gated). Paper: [DeepSpot-M](https://www.medrxiv.org/content/10.64898/2026.06.19.26356060v1).
 - [Therapeutics Data Commons (TDC)](https://tdcommons.ai/) — Unified benchmark suite covering ADMET, drug-target interaction, drug response, and more.
 - [Tox21](https://tripod.nih.gov/tox21/challenge/) — 12,707 compounds tested in 12 nuclear receptor and stress-response pathway biochemical assays for toxicity prediction.
 ### Genomics, Cancer & Biomedical Cohorts
 - [1000 Genomes Project](https://www.internationalgenome.org/) — Reference panel of human genetic variation from 2,504 individuals across 26 populations.
 - [CPTAC (Clinical Proteomic Tumor Analysis Consortium)](https://proteomics.cancer.gov/programs/cptac) — Multi-omic proteogenomic datasets for multiple cancer types linking proteomics with genomics.
 - [The Cancer Genome Atlas (TCGA)](https://www.cancer.gov/about-nci/organization/ccg/research/structural-genomics/tcga) — Comprehensive multi-omics (genomics, transcriptomics, proteomics, methylation) dataset for 33 cancer types across ~11,000 patients.
 - [UK Biobank](https://www.ukbiobank.ac.uk/) — Large-scale biomedical database of ~500K participants with genetic, imaging, and health data for population genetics and disease studies.
 ### General Biological ML & Knowledge Graph Benchmarks
 - [OGB (Open Graph Benchmark)](https://ogb.stanford.edu/) — Large-scale graph ML benchmark suite including biological datasets such as ogbl-ppa (protein-protein associations) and ogbg-molhiv.
 - [OpenBioLink](https://github.com/OpenBioLink/OpenBioLink) — Benchmark datasets for biological knowledge graph completion.
 ---
 ## API
@@ -344,6 +398,11 @@ Browse and search the resources via the [GitHub Pages UI](https://inoue0426.gith
 - [TGSA](https://github.com/violet-sto/TGSA) — Tumor gene set and attention-based model leveraging biological pathway knowledge for drug response prediction.
 - [HiDRA](https://github.com/bsml320/HiDRA) — Hierarchical network model incorporating gene and pathway-level information for cancer drug response prediction.
 - [DRUML](https://github.com/CutillasLab/DRUMLR) — Ensemble machine learning framework combining standard ML with deep learning to systematically rank anti-cancer drugs from proteomics and RNA-seq data.
 - [PASO](https://github.com/queryang/PASO) — Pathway-aware multi-omics drug response model combining pathway-difference features, multi-scale convolutions, Transformer encoding, and drug SMILES.
 - [DTLCDR](https://doi.org/10.1016/j.jpha.2025.101315) — Target-based multimodal framework for preclinical cancer drug response prediction and transfer to clinical response, with explicit unseen-drug generalization.
 - [THERAPI](https://github.com/Sunginyoung/THERAPI) — Cell-line-to-patient transfer framework that aligns tumor transcriptomes with cancer cell lines and integrates perturbation and gene-level representations for patient drug response prediction.
 - [EXPRESSO](https://doi.org/10.1158/0008-5472.CAN-25-5220) — Supervised treatment-response framework using pretreatment tumor transcriptomics, drug targets, and context-specific biomarkers across multiple cancer types and therapies.
 - [PerturbRx](https://arxiv.org/abs/2608.21349) — Treatment-conditioned representation learning framework that transfers drug-induced latent transitions learned from single-cell perturbation data to patient-level cancer treatment-response prediction.
 #### Drug Perturbation
@@ -352,6 +411,16 @@ Browse and search the resources via the [GitHub Pages UI](https://inoue0426.gith
 - [chemCPA](https://github.com/theislab/chemCPA) — Compositional perturbation autoencoder for predicting single-cell transcriptional responses to unseen drug perturbations and dose combinations.
 - [cycleCDR](https://github.com/hliulab/cycleCDR) — Interpretable cycle-consistency framework for modeling cellular responses to drug perturbations.
 - [PRNet](https://github.com/Perturbation-Response-Prediction/PRnet) — Deep generative model for predicting transcriptional responses to novel chemical perturbations for drug discovery.
 - [Bison](https://arxiv.org/abs/2609.32467) — Cross-dataset model for globally unseen-compound response prediction using a shared gene representation, discrete diffusion models, and matched drug-contrast supervision.
 - [biolord](https://github.com/nitzanlab/biolord) — Deep generative model that disentangles known and unknown attributes for conditional generation of single-cell states.
 - [PerturbNet](https://github.com/welch-lab/PerturbNet) — Conditional generative model for predicting distributions of single-cell states under unseen chemical and genetic perturbations.
 - [CellFlow](https://github.com/theislab/CellFlow) — Conditional flow-matching framework for modeling and predicting cellular phenotypes under chemical, genetic, and other perturbations.
 - [Prophet](https://github.com/theislab/prophet) — Transformer model for predicting cellular phenotypes under unseen chemical or genetic perturbations across heterogeneous assays and contexts.
 - [PrePR-CT](https://github.com/reem12345/Cell-Type-Specific-Graphs) — Graph-based model using cell-type-specific co-expression networks as inductive priors for small-data chemical perturbation response prediction.
 - [XPert](https://github.com/GSanShui/XPert) — Knowledge-informed dual-branch Transformer for drug-induced transcriptional perturbation prediction across dose, time, and cellular context.
 - [State](https://github.com/ArcInstitute/state) — Transition model for predicting cellular perturbation responses across diverse contexts and sets of cells.
 - [TxPert](https://github.com/valence-labs/TxPert) — Knowledge-graph-informed latent-transfer model for transcriptomic perturbation prediction across unseen single perturbations, combinations, and cross-context settings.
 - [LPM](https://github.com/perturblib/perturblib) — Large perturbation model that jointly learns heterogeneous perturbation experiments by disentangling perturbation, readout, and context representations.
 #### Drug Repurposing
@@ -396,11 +465,45 @@ Browse and search the resources via the [GitHub Pages UI](https://inoue0426.gith
 - [GeneGPT](https://github.com/ncbi/GeneGPT) — LLM for biomedical information, integrated with various APIs.
 - [GenePT](https://github.com/yiqunchen/GenePT) — Foundation LLM for single-cell data.
 - [scPRINT](https://github.com/cantinilab/scPRINT) — Pretrained on 50M cells for scRNA-seq denoising & zero imputation.
 - [ClawBio](https://github.com/ClawBio/ClawBio) — Bioinformatics-native AI agent skill library with local-first pharmacogenomics, ancestry PCA, semantic similarity, nutrigenomics, and metagenomics skills.
 - [BioMedLM](https://huggingface.co/stanford-crfm/BioMedLM) — 2.7B parameter GPT-2-style language model trained exclusively on biomedical literature from PubMed for biomedical question answering and text generation.
 - [MolT5](https://github.com/blender-nlp/MolT5) — Language model for molecular tasks bridging text and SMILES, enabling molecule captioning and text-driven molecule generation.
 - [ChatDrug](https://github.com/chao1224/ChatDrug) — LLM-based conversational pipeline for drug discovery, using natural language prompts for iterative drug editing and optimization.
- [CASSIA](https://github.com/ElliotXie/CASSIA) — Multi-agent LLM for reference-free, interpretable cell-type annotation of single-cell RNA-seq data, with dedicated annotation, validation, scoring, and reporting agents.
+
 ### Agentic AI for Biology
 #### General Biomedical Agents
 - [Biomni](https://github.com/snap-stanford/Biomni) — General-purpose biomedical AI agent integrating planning, code execution, specialized tools, databases, and software across diverse biomedical research tasks.
 - [ToolUniverse](https://github.com/mims-harvard/ToolUniverse) — Unified scientific tool ecosystem for building AI scientists that can discover, select, and execute biomedical tools and databases.
 - [ClawBio](https://github.com/ClawBio/ClawBio) — Bioinformatics-native AI agent skill library with local-first pharmacogenomics, ancestry PCA, semantic similarity, nutrigenomics, and metagenomics skills.
 - [BioMedAgent](https://github.com/BOBQWERA/BioMedAgent) — Self-evolving multi-agent framework for autonomous biomedical data analysis with tool discovery, workflow planning, code generation, execution, correction, and cross-omics analysis.
 - [BioMaster](https://github.com/ai4nucleome/BioMaster) — Multi-agent system for automated and auditable bioinformatics workflows spanning RNA-seq, ChIP-seq, single-cell, spatial omics, Hi-C, long reads, metagenomics, and proteomics.
 - [BRAD](https://github.com/Jpickard1/BRAD) — Retrieval-augmented bioinformatics assistant integrating scientific literature, databases, external tools, and executable workflows.
 #### Therapeutics & Drug Discovery Agents
 - [TxAgent](https://github.com/mims-harvard/TxAgent) — Therapeutic reasoning agent using multi-step reasoning and a large scientific tool universe for drug interactions, contraindications, and personalized treatment analysis.
 - [Medea](https://github.com/mims-harvard/Medea) — Multi-agent therapeutic discovery system combining research planning, biological data analysis, literature reasoning, and multi-LLM deliberation across single-cell, cell-line, and patient contexts.
 - [DrugAgent](https://github.com/inoue0426/DrugAgent) — Multi-agent biomedical evidence synthesis framework for computational drug discovery with reliability-aware aggregation.
 #### Bioinformatics & Omics Agents
 - [CASSIA](https://github.com/ElliotXie/CASSIA) — Multi-agent LLM framework for reference-free and interpretable single-cell cell-type annotation with dedicated annotation, validation, scoring, and reporting agents.
 - [STELLA](https://github.com/zaixizhang/STELLA) — Self-evolving biomedical research agent that expands its tool repertoire and supports literature reasoning, computational analysis, and laboratory-oriented scientific workflows.
 - [AutoBA](https://github.com/JoshuaChou2018/AutoBA) — Automated multi-omics analysis agent that plans, executes, and repairs bioinformatics workflows from natural-language objectives.
 - [GenoMAS](https://github.com/Liu-Hy/GenoMAS) — Multi-agent framework for code-driven gene-expression analysis with planning, execution, debugging, backtracking, and GEO/TCGA-based scientific discovery.
 - [BIA](https://github.com/biagent-dev/bia) — Bioinformatics agent for GEO search, sample metadata extraction, count-matrix processing, and pipeline extraction from papers.
 - [BioAgents](https://github.com/microsoft/bioinformagus) — Multi-agent bioinformatics assistant using specialized language models and retrieval for genomics workflow development and troubleshooting.
 - [Genomi](https://github.com/exon-research/genomi) — Local-first genomics agent runtime that indexes personal variants, queries evidence, and generates evidence-grounded reports while keeping raw genome data on-device.
 #### Multi-Agent Scientific Labs
 - [Virtual Lab](https://github.com/zou-group/virtual-lab) — Human–AI collaborative research environment in which an LLM principal investigator coordinates specialized scientist agents for scientific discovery.
 - [Agent Laboratory](https://github.com/SamuelSchmidgall/AgentLaboratory) — End-to-end multi-agent research workflow for literature review, experimentation, implementation, analysis, and report generation.
 #### Paper & Workflow Agents
 - [Paper2Agent](https://github.com/jmiao24/Paper2Agent) — Multi-agent system that transforms research papers and associated code into interactive, testable scientific agents and MCP tools.
 ### Foundation Models
@@ -418,6 +521,8 @@ Browse and search the resources via the [GitHub Pages UI](https://inoue0426.gith
 - [GEARS](https://github.com/snap-stanford/GEARS) — Graph-based model for predicting transcriptional responses to single and combinatorial genetic perturbations using biological priors.
 - [SATURN](https://github.com/snap-stanford/SATURN) — Transformer-based model integrating gene expression and protein sequences via a protein language model to learn unified multi-species cell embeddings.
 - [CancerFoundation](https://github.com/BoevaLab/CancerFoundation) — Single-cell RNA-seq foundation model trained exclusively on a curated dataset of malignant cells to learn cancer-specific embeddings.
 - [CellFM](https://github.com/biomed-AI/CellFM) — 800M-parameter single-cell foundation model pretrained on transcriptomics from 100 million human cells for annotation, integration, gene-function, and perturbation tasks.
 - [scPRINT-2](https://github.com/cantinilab/scPRINT-2) — Next-generation single-cell foundation model pretrained on 350M+ cells across 22K+ datasets and 16 species for embeddings, denoising, annotation, gene-network inference, and cross-species transfer.
 ##### Spatial Foundation Models
@@ -432,6 +537,16 @@ Browse and search the resources via the [GitHub Pages UI](https://inoue0426.gith
 - [DeepSpot-M](https://github.com/ratschlab/DeepSpotM) — Multimodal foundation model for transcriptome-wide virtual spatial transcriptomics from histology.
 - [AESTETIK](https://github.com/ratschlab/aestetik) — Autoencoder for spatial transcriptomics representation learning using topology and histology image knowledge.
 ##### Pathology Foundation Models
 - [TITAN](https://github.com/mahmoodlab/TITAN) — Multimodal whole-slide pathology foundation model that combines image and language supervision for slide-level representation and zero-shot analysis.
 - [Virchow2](https://huggingface.co/paige-ai/Virchow2) — 632M-parameter pathology vision transformer pretrained on 3.1M whole-slide images with mixed-magnification self-supervision.
 - [H-Optimus-0](https://huggingface.co/bioptimus/H-optimus-0) — 1.1B-parameter histopathology foundation model trained with self-supervised learning on a large multi-center slide corpus.
 - [H-Optimus-1](https://huggingface.co/bioptimus/H-optimus-1) — 1.1B-parameter pathology foundation model trained on billions of histology images from more than one million slides and 800K+ patients.
 - [UNI2-h](https://huggingface.co/MahmoodLab/UNI2-h) — Billion-parameter histopathology vision foundation model for tile-level feature extraction and downstream computational pathology tasks.
 - [Phikon-v2](https://huggingface.co/owkin/phikon-v2) — Updated pathology foundation model for general-purpose histology feature extraction and transfer learning.
 - [GenBio-PathFM](https://github.com/genbio-ai/genbio-pathfm) — 1.1B-parameter histopathology foundation model trained on public data using morphology-aware curation and dual-stage JEPA+DINO learning.
 ##### Multi-Omics Foundation Models
 - [scMulan](https://github.com/SuperBianC/scMulan) — Single-cell multi-omic language model pretrained on ~10M cells spanning transcriptomics, epigenomics, and proteomics for cross-omics transfer tasks.
@@ -465,6 +580,8 @@ Browse and search the resources via the [GitHub Pages UI](https://inoue0426.gith
 - [Mol2Vec](https://github.com/samoturk/mol2vec) — Unsupervised molecular embedding method inspired by Word2Vec for learning vector representations of chemical substructures.
 - [MolFormer](https://github.com/IBM/molformer) — Linear attention transformer pretrained on millions of SMILES strings for efficient molecular embeddings.
 - [Uni-Mol](https://github.com/deepmodeling/Uni-Mol) — 3D molecular pretraining framework for universal representation learning on molecules and protein pockets.
 - [Uni-Mol2](https://github.com/deepmodeling/Uni-Mol/tree/main/unimol2) — Scaled molecular pretraining model using atomic, graph, and 3D geometry features, with models up to 1.1B parameters pretrained on 800M conformations.
 - [ChemFM](https://github.com/TheLuoFengLab/ChemFM) — 1B/3B-parameter chemical language model pretrained on 178M molecules for molecular representation, property prediction, generation, and synthesis tasks.
 #### Protein Foundation Models
@@ -474,11 +591,14 @@ Browse and search the resources via the [GitHub Pages UI](https://inoue0426.gith
 - [ProtTrans](https://github.com/agemagician/ProtTrans) — Suite of protein language models (ProtBERT, ProtT5, ProtXLNet) trained on billions of protein sequences from UniRef and BFD.
 - [ProGen2](https://github.com/salesforce/progen) — Protein language model trained on diverse protein families for sequence generation and fitness prediction.
 - [Ankh](https://github.com/agemagician/Ankh) — Efficient protein language model optimized for downstream prediction tasks including secondary structure, localization, and function annotation.
 - [ESM Cambrian (ESM C)](https://github.com/Biohub/esm) — Protein representation foundation-model family designed as an efficient next-generation successor to ESM2, spanning 300M to multi-billion-parameter models.
 ##### Protein Structure Prediction and Design
 - [AlphaFold3](https://github.com/google-deepmind/alphafold3) — Predicts structures of proteins, nucleic acids, small molecules, and their complexes.
 - [Boltz-1](https://github.com/jwohlwend/boltz) — Open-source all-atom biomolecular structure prediction model for proteins, nucleic acids, small molecules, and their complexes achieving AlphaFold3-level accuracy.
 - [Boltz-2](https://github.com/jwohlwend/boltz) — Biomolecular foundation model jointly predicting complex structures and binding affinities for protein–ligand interaction modeling and virtual screening.
 - [Protenix](https://github.com/bytedance/Protenix) — Trainable biomolecular structure-prediction framework for proteins, nucleic acids, ligands, and complexes with open training and inference pipelines.
 - [Chai-1](https://github.com/chaidiscovery/chai-lab) — Unified molecular structure prediction model covering proteins, nucleic acids, small molecules, and complexes.
 - [ESM3](https://github.com/evolutionaryscale/esm) — Multimodal protein language model that jointly reasons over sequence, structure, and function for generative protein design and engineering.
 - [ESMFold](https://github.com/facebookresearch/esm) — Fast protein structure prediction using language model embeddings.
@@ -501,6 +621,36 @@ Browse and search the resources via the [GitHub Pages UI](https://inoue0426.gith
 - [PLIP](https://github.com/PathologyFoundation/plip) — Vision-language foundation model for pathology trained with contrastive learning on pathology image–text pairs for image classification and text-to-image retrieval.
 - [MUSK](https://github.com/lilab-stanford/MUSK) — Vision-language foundation model for precision oncology analyzing multimodal paired text and pathology image data for biomarker prediction and retrieval.
 #### Cancer Genome Foundation Models
 - [TESSERA](https://github.com/JW-Sidhom-Lab/tessera) — Cancer-genome foundation model jointly pretrained on somatic SNVs and copy-number alterations from TCGA using masked reconstruction and cross-modal contrastive learning.
 - [MutationProjector](https://doi.org/10.1101/2025.09.08.674723) — Pan-cancer genotype foundation model trained on mutations and copy-number alterations from >30K tumors for clinical representation learning.
 #### Single-Cell Epigenomics Foundation Models
 - [EpiAgent](https://github.com/xy-chen16/EpiAgent) — scATAC-seq foundation model pretrained on ~5M cells and >35B tokens for representation learning, annotation, imputation, perturbation prediction, and in-silico cCRE knockout.
 - [SCARF](https://doi.org/10.1101/2025.04.07.647689) — Single-cell RNA+ATAC foundation model pretrained on >2.7M cells for multimodal representation, matching, cross-omics translation, and few-shot annotation.
 - [scDNAm-GPT](https://github.com/ChaoqiLiang/scDNAm-GPM) — Foundation model for single-cell whole-genome bisulfite sequencing with whole-genome context modeling at single-CpG resolution.
 - [EpiFoundation](https://doi.org/10.1101/2025.02.05.636688) — Foundation model for scATAC-seq using peak-to-gene aligned pretraining for cell representation, annotation, batch correction, and gene-expression prediction.
 - [ChromFound](https://github.com/SAIS-LifeScience/ChromFound) — Genome-aware scATAC-seq foundation model pretrained on 1.97M cells across tissues and disease contexts for zero-shot cell representations, annotation, and cross-omics prediction.
 - [Atacformer](https://doi.org/10.1101/2025.11.03.685753) — Transformer foundation model for scATAC-seq that learns embeddings of cis-regulatory elements for clustering, annotation, and reference mapping.
 - [CLM-X](https://doi.org/10.64898/2026.02.17.704943) — Multi-way Transformer foundation model jointly handling RNA-only, ATAC-only, and paired RNA–ATAC single-cell inputs for integration, translation, annotation, and perturbation prediction.
 #### Other Omics Foundation Models
 - [CAPTAIN](https://doi.org/10.1038/s41467-026-72882-y) — Multimodal foundation model pretrained on co-assayed single-cell RNA and protein for joint representation learning and cross-modal downstream tasks.
 - [HiCFoundation](https://doi.org/10.1038/s41592-026-03097-8) — Hi-C foundation model pretrained on large-scale chromatin-contact maps for 3D-genome analysis, epigenomic prediction, and single-cell Hi-C adaptation.
 - [OmicsFM](https://github.com/CompOmics/OmicsFM) — Modality-agnostic molecular-expression foundation model with matched proteomics, bulk-transcriptomics, and single-cell-transcriptomics checkpoints.
 - [VirTues](https://github.com/bunnelab/virtues) — Spatial-proteomics foundation model learning marker-aware representations across proteins, cells, niches, and tissues from multiplexed imaging.
 - [MethylGPT](https://github.com/albert-ying/MethylGPT) — Transformer foundation model for DNA methylation pretrained on >150K human methylomes across thousands of datasets, with 3M/7M/15M parameter variants.
 - [CpGPT](https://github.com/lucascamillomd/CpGPT) — DNA methylation foundation model pretrained on >150K samples for zero-shot imputation, array conversion, reference mapping, and downstream phenotype prediction.
 - [Casanovo Foundation](https://github.com/Noble-Lab/casanovo-tl) — Tandem mass-spectrometry proteomics foundation model that reuses a pretrained Casanovo spectrum encoder for spectrum quality, chimericity, and post-translational-modification prediction.
 #### RNA Foundation Models
 - [RNA-FM](https://github.com/ml4bio/RNA-FM) — General-purpose RNA foundation model pretrained on large-scale RNA sequences for structural and functional representation learning.
 - [RiNALMo](https://github.com/lbcb-sci/RiNALMo) — RNA language-model family pretrained on tens of millions of RNA sequences for secondary-structure and functional prediction tasks.
 #### Genomics Foundation Models
 - [Nucleotide Transformer](https://github.com/instadeepai/nucleotide-transformer) — Foundation model for genomic sequences across multiple species.
@@ -510,6 +660,9 @@ Browse and search the resources via the [GitHub Pages UI](https://inoue0426.gith
 - [Basenji](https://github.com/calico/basenji) — Sequential regulatory activity prediction from DNA sequences.
 - [Caduceus](https://github.com/kuleshov-group/caduceus) — Bidirectional equivariant long-range DNA sequence model based on Mamba.
 - [Evo](https://github.com/evo-design/evo) — Long-context genomic foundation model (up to 1M tokens).
 - [Evo 2](https://github.com/arcinstitute/evo2) — Genome foundation model trained on 9 trillion DNA base pairs across all domains of life with a 1M-token context window and single-nucleotide resolution.
 - [AlphaGenome](https://github.com/google-deepmind/alphagenome) — Long-context DNA model predicting multimodal regulatory outputs including expression, splicing, chromatin features, and contact maps at near base-pair resolution.
 - [modernGENA](https://github.com/AIRI-Institute/GENA_LM) — ModernBERT-style DNA foundation-model family pretrained on hundreds of vertebrate genome assemblies for efficient long-sequence regulatory modeling.
 - [HyenaDNA](https://github.com/HazyResearch/hyena-dna) — Long-range genomic foundation model handling sequences up to 1M tokens with sub-quadratic attention.
 - [Borzoi](https://github.com/calico/borzoi) — Extended successor to Enformer for predicting RNA-seq coverage from long genomic sequence windows (524 kb) with improved resolution.
 - [DeepSEA](http://deepsea.princeton.edu/) — Deep learning framework for predicting chromatin effects of sequence alterations with single-nucleotide sensitivity across thousands of chromatin features.
@@ -2,9 +2,9 @@
 title: "Cspell"
 task: ""
 lineage_type: import
-upstream_source: https://github.com/inoue0426/awesome-computational-biology/blob/c6f07d90/cspell.json
+upstream_source: https://github.com/inoue0426/awesome-computational-biology/blob/4ed7cfef/cspell.json
-upstream_sha: c6f07d90
+upstream_sha: 4ed7cfef
-imported_at: 2026-08-31
+imported_at: 2026-10-26
 prompt_class: unknown
 upstream_changes: accepted
 author: upstream
@@ -15,166 +15,249 @@ validated: false
  "language": "en",
  "allowCompoundWords": true,
  "words": [
-    "behavioural",
+    "Adamson",
-    "KEGG",
+    "ADME",
    "NCBI",
    "UCSC",
    "EMBL",
    "RCSB",
    "CASP",
    "Uniclust",
    "Reactome",
    "Bioactive",
    "biopolymers",
    "proteomics",
    "transcriptomics",
    "metagenomic",
    "metatranscriptomic",
    "CRISPR",
    "JASPAR",
    "druggable",
    "Toxicogenomics",
    "GDSC",
    "biomolecular",
    "DRKG",
    "Hetionet",
    "Eudra",
    "esearch",
    "efetch",
    "Ensembl",
    "Cheminformatics",
    "Deconv",
    "Scanpy",
    "Squidpy",
    "explainability",
    "MOFGCN",
    "Autoencoder",
    "DGDRP",
    "MCPINN",
    "Pretrained",
    "pretrained",
    "denoising",
    "transcriptomic",
    "CELLxGENE",
    "eukaryotic",
    "metabolites",
    "OMIM",
    "OncoKB",
    "Mendelian",
    "DisGeNET",
    "GWAS",
    "IntAct",
    "Biopython",
    "MDAnalysis",
    "trajectories",
    "Geneformer",
    "equivariant",
    "HyenaDNA",
    "Hyena",
    "Caduceus",
    "Mamba",
    "retrosynthesis",
    "TargetDiff",
    "Chai",
    "Zuckerberg",
    "HMDB",
    "CTRP",
    "ADMET",
-    "Omics",
+    "AESTETIK",
-    "omics",
+    "AlphaGenome",
-    "omic",
+    "Ankh",
-    "OADB",
+    "ATAC",
    "Atacformer",
    "AutoBA",
    "Autoencoder",
    "BACE",
    "behavioural",
    "Bento",
    "Bioactive",
    "BioAgents",
    "biolord",
    "BioMaster",
    "BioMedAgent",
    "Biomni",
    "biomolecular",
    "biopolymers",
    "Biopython",
    "Bison",
    "Boltz",
    "bowang",
    "Caduceus",
    "Cambrian",
    "Casanovo",
    "CASP",
    "CASSIA",
    "CellFlow",
    "CellFM",
    "CellOT",
    "CELLxGENE",
    "Chai",
    "ChatDrug",
    "Chem",
    "ChemFM",
    "Cheminformatics",
    "chimericity",
    "Chrom",
    "ChromFound",
    "CITE-seq",
    "ClawBio",
    "Clin",
    "CMonge",
    "convolutional",
    "CpGPT",
    "CPTAC",
    "CRISPR",
    "CRISPRa",
    "CRISPRi",
    "CROP",
    "ctheodoris",
    "CTRP",
    "Datlinger",
    "Deconv",
    "deconvolution",
    "DeepSEA",
    "denoising",
    "DGDRP",
    "DisGeNET",
    "Dixit",
    "DoubletFinder",
    "DRKG",
    "druggable",
    "DRUML",
    "DTLCDR",
    "efetch",
    "EMBL",
    "Enformer",
    "Ensembl",
    "EpiFoundation",
    "epigenome",
    "epigenomic",
    "epigenomics",
    "equivariant",
    "esearch",
    "Eudra",
    "eukaryotic",
    "explainability",
    "EXPRESSO",
    "FANTOM",
    "FCFP4",
    "FFPE",
    "finetunable",
    "Frangieh",
    "Gasperini",
    "GDSC",
    "GenBio",
    "Geneformer",
    "Geno",
    "GenoMAS",
    "Genomi",
    "Giga",
    "Gnify",
    "gnom",
-    "Rfam",
+    "GPN",
    "GPT",
    "Guaca",
-    "deconvolution",
+    "GWAS",
-    "scvi",
+    "H-Optimus",
-    "pharmacogenomics",
+    "Haplotype",
-    "nutrigenomics",
+    "Hetionet",
-    "Giga",
+    "HiDRA",
-    "Phikon",
+    "HMDB",
-    "TCGA",
+    "Hyena",
-    "Mulan",
+    "HyenaDNA",
-    "epigenomics",
+    "Imageomics",
-    "methylation",
+    "IntAct",
-    "ATAC",
+    "JASPAR",
-    "MOFA",
+    "JEPA",
    "TOSICA",
    "Boltz",
    "MPNN",
    "Enformer",
    "Velo",
    "BACE",
    "secretase",
    "Clin",
    "CPTAC",
    "Proteomic",
    "proteogenomic",
    "LINCS",
    "ogbl",
    "ogbg",
    "SIDER",
    "Muris",
    "Pfam",
    "PROSITE",
    "epigenome",
    "TRRUST",
    "kallisto",
    "pseudoalignment",
    "multiplet",
    "TGSA",
    "JTVAE",
    "Jurkat",
    "kallisto",
    "KEGG",
    "KIBA",
    "L1000",
    "ligand",
    "LINCS",
    "Mamba",
    "MCPINN",
    "MDAnalysis",
    "Medea",
    "Mendelian",
    "metabolites",
    "metagenomic",
    "metatranscriptomic",
    "methylation",
    "MethylGPT",
    "methylomes",
    "miRBase",
    "miRNA",
-    "ProtTrans",
+    "modernGENA",
-    "ProtBERT",
+    "MOFA",
-    "ProGen",
+    "MOFGCN",
    "Ankh",
    "DeepSEA",
    "RegNetwork",
    "ROADMAP",
    "FANTOM",
    "NeXtProt",
    "HiDRA",
    "MolT",
    "ChatDrug",
    "DoubletFinder",
    "pseudotime",
    "ligand",
    "SCENIC",
    "GPN",
    "Sei",
    "KIBA",
    "pharmacokinetics",
    "ADME",
    "Haplotype",
    "NCEM",
    "multiome",
    "MOGONET",
-    "convolutional",
+    "MolT",
-    "DRUML",
+    "Morgan",
-    "SBDD",
+    "MPNN",
-    "Pacc",
+    "Mulan",
    "multiome",
    "multiomics",
-    "Pathomic",
+    "multiplet",
-    "PLIP",
+    "Muris",
-    "Omni",
+    "MutationProjector",
    "Bento",
    "FFPE",
    "Xenium",
    "Zyme",
    "Neur",
    "Imageomics",
    "AESTETIK",
    "CellOT",
    "CMonge",
    "bowang",
    "ctheodoris",
    "OpenAI",
    "GPT",
    "nanobody",
-    "thermostability"
+    "NCBI",
    "NCEM",
    "Neur",
    "NeXtProt",
    "Novartis",
    "nutrigenomics",
    "OADB",
    "ogbg",
    "ogbl",
    "omic",
    "omics",
    "Omics",
    "OmicsFM",
    "OMIM",
    "Omni",
    "OncoKB",
    "OP3",
    "OpenAI",
    "Pacc",
    "Paper2Agent",
    "Pathomic",
    "Perturb",
    "Perturb-seq",
    "PerturbNet",
    "PerturBridge",
    "PerturbRx",
    "Pfam",
    "pharmacogenomics",
    "pharmacokinetics",
    "Phikon",
    "PLIP",
    "PrePR",
    "pretrained",
    "Pretrained",
    "ProGen",
    "PROSITE",
    "ProtBERT",
    "Protenix",
    "proteogenomic",
    "Proteomic",
    "proteomics",
    "ProtTrans",
    "pseudoalignment",
    "pseudobulk",
    "pseudotime",
    "RCSB",
    "RDKit",
    "Reactome",
    "RegNetwork",
    "Replogle",
    "retrosynthesis",
    "Rfam",
    "RiNALMo",
    "ROADMAP",
    "SBDD",
    "Scanpy",
    "SCENIC",
    "SciPlex",
    "scPRINT",
    "scrna",
    "scvi",
    "secretase",
    "Sei",
    "SIDER",
    "silico",
    "Squidpy",
    "State",
    "STELLA",
    "TargetDiff",
    "TCGA",
    "TGSA",
    "THERAPI",
    "thermostability",
    "TITAN",
    "ToolUniverse",
    "TOSICA",
    "Toxicogenomics",
    "trajectories",
    "transcriptomic",
    "transcriptomics",
    "TRRUST",
    "TxAgent",
    "TxPert",
    "UCSC",
    "Uni-Mol2",
    "UNI2",
    "Uniclust",
    "UniMol",
    "VCPI",
    "Velo",
    "Virchow2",
    "VirTues",
    "Xenium",
    "XPert",
    "Zuckerberg",
    "Zyme"
  ],
  "ignorePaths": [
    "node_modules/**"
@@ -0,0 +1,298 @@
 ---
 title: "Agent Explorer v1 metadata."
 task: ""
 lineage_type: import
 upstream_source: https://github.com/inoue0426/awesome-computational-biology/blob/4ed7cfef/data/enrichment.agent-explorer-v1.yml
 upstream_sha: 4ed7cfef
 imported_at: 2026-10-26
 prompt_class: unknown
 upstream_changes: accepted
 author: upstream
 validated: false
 ---
 # Agent Explorer v1 metadata.
 resources:
  biomni:
    agent_profile:
      year: 2025
      domains: [general-biomedicine, bioinformatics, therapeutics, omics]
      architecture: single-agent
      tool_use: true
      code_execution: true
      web_retrieval: true
      literature: true
      omics: true
      wet_lab: limited
      autonomous_experiment: limited
      human_in_loop: true
      open_source: true
  tooluniverse:
    agent_profile:
      year: 2025
      domains: [general-biomedicine, tool-ecosystem, therapeutics, bioinformatics]
      architecture: agent-ecosystem
      tool_use: true
      code_execution: limited
      web_retrieval: true
      literature: true
      omics: true
      wet_lab: limited
      autonomous_experiment: limited
      human_in_loop: true
      open_source: true
  clawbio:
    agent_profile:
      year: 2026
      domains: [bioinformatics, genomics, pharmacogenomics, omics]
      architecture: agent-ecosystem
      tool_use: true
      code_execution: true
      web_retrieval: true
      literature: limited
      omics: true
      wet_lab: false
      autonomous_experiment: false
      human_in_loop: true
      open_source: true
  txagent:
    agent_profile:
      year: 2025
      domains: [therapeutics, drug-discovery, precision-medicine]
      architecture: single-agent
      tool_use: true
      code_execution: limited
      web_retrieval: true
      literature: true
      omics: limited
      wet_lab: false
      autonomous_experiment: limited
      human_in_loop: true
      open_source: true
  medea:
    agent_profile:
      year: 2025
      domains: [therapeutics, drug-discovery, single-cell, omics]
      architecture: multi-agent
      tool_use: true
      code_execution: true
      web_retrieval: true
      literature: true
      omics: true
      wet_lab: limited
      autonomous_experiment: limited
      human_in_loop: true
      open_source: true
  drugagent:
    agent_profile:
      year: 2026
      domains: [drug-discovery, therapeutics, evidence-synthesis]
      architecture: multi-agent
      tool_use: true
      code_execution: limited
      web_retrieval: true
      literature: true
      omics: limited
      wet_lab: false
      autonomous_experiment: limited
      human_in_loop: true
      open_source: true
  cassia:
    agent_profile:
      year: 2024
      domains: [single-cell, bioinformatics, omics]
      architecture: multi-agent
      tool_use: limited
      code_execution: limited
      web_retrieval: limited
      literature: true
      omics: true
      wet_lab: false
      autonomous_experiment: false
      human_in_loop: true
      open_source: true
  stella:
    agent_profile:
      year: 2025
      domains: [general-biomedicine, bioinformatics, scientific-workflow]
      architecture: single-agent
      tool_use: true
      code_execution: true
      web_retrieval: true
      literature: true
      omics: limited
      wet_lab: limited
      autonomous_experiment: true
      human_in_loop: true
      open_source: true
  virtual_lab:
    agent_profile:
      year: 2024
      domains: [scientific-workflow, protein-design, experimental-design]
      architecture: multi-agent
      tool_use: true
      code_execution: true
      web_retrieval: limited
      literature: true
      omics: limited
      wet_lab: true
      autonomous_experiment: limited
      human_in_loop: true
      open_source: true
  agent_laboratory:
    agent_profile:
      year: 2025
      domains: [scientific-workflow, research-automation]
      architecture: multi-agent
      tool_use: true
      code_execution: true
      web_retrieval: true
      literature: true
      omics: false
      wet_lab: false
      autonomous_experiment: true
      human_in_loop: true
      open_source: true
  paper2agent:
    agent_profile:
      year: 2025
      domains: [paper-to-agent, scientific-workflow, tool-generation]
      architecture: multi-agent
      tool_use: true
      code_execution: true
      web_retrieval: true
      literature: true
      omics: limited
      wet_lab: false
      autonomous_experiment: limited
      human_in_loop: true
      open_source: true
  biomedagent:
    agent_profile:
      year: 2026
      domains: [general-biomedicine, bioinformatics, omics]
      architecture: multi-agent
      tool_use: true
      code_execution: true
      web_retrieval: limited
      literature: limited
      omics: true
      wet_lab: false
      autonomous_experiment: true
      human_in_loop: true
      open_source: true
  biomaster:
    agent_profile:
      year: 2026
      domains: [bioinformatics, omics, scientific-workflow]
      architecture: multi-agent
      tool_use: true
      code_execution: true
      web_retrieval: limited
      literature: limited
      omics: true
      wet_lab: false
      autonomous_experiment: true
      human_in_loop: true
      open_source: true
  brad:
    agent_profile:
      year: 2025
      domains: [bioinformatics, literature, scientific-workflow]
      architecture: single-agent
      tool_use: true
      code_execution: true
      web_retrieval: true
      literature: true
      omics: limited
      wet_lab: false
      autonomous_experiment: limited
      human_in_loop: true
      open_source: true
  autoba:
    agent_profile:
      year: 2024
      domains: [bioinformatics, omics, scientific-workflow]
      architecture: single-agent
      tool_use: true
      code_execution: true
      web_retrieval: limited
      literature: limited
      omics: true
      wet_lab: false
      autonomous_experiment: true
      human_in_loop: true
      open_source: true
  genomas:
    agent_profile:
      year: 2025
      domains: [bioinformatics, gene-expression, scientific-workflow]
      architecture: multi-agent
      tool_use: true
      code_execution: true
      web_retrieval: limited
      literature: true
      omics: true
      wet_lab: false
      autonomous_experiment: true
      human_in_loop: true
      open_source: true
  bia:
    agent_profile:
      domains: [bioinformatics, genomics, literature]
      architecture: single-agent
      tool_use: true
      code_execution: true
      web_retrieval: true
      literature: true
      omics: true
      wet_lab: false
      autonomous_experiment: limited
      human_in_loop: true
      open_source: true
  bioagents:
    agent_profile:
      year: 2025
      domains: [bioinformatics, genomics, scientific-workflow]
      architecture: multi-agent
      tool_use: true
      code_execution: limited
      web_retrieval: limited
      literature: limited
      omics: true
      wet_lab: false
      autonomous_experiment: limited
      human_in_loop: true
      open_source: true
  genomi:
    agent_profile:
      year: 2026
      domains: [genomics, precision-medicine, evidence-synthesis]
      architecture: agent-ecosystem
      tool_use: true
      code_execution: limited
      web_retrieval: limited
      literature: true
      omics: limited
      wet_lab: false
      autonomous_experiment: false
      human_in_loop: true
      open_source: true
@@ -2,9 +2,9 @@
 title: "Provenance-backed single-cell and biomedical benchmark enrichment batch."
 task: ""
 lineage_type: import
-upstream_source: https://github.com/inoue0426/awesome-computational-biology/blob/7a064bf0/data/enrichment.benchmark-v2.yml
+upstream_source: https://github.com/inoue0426/awesome-computational-biology/blob/4ed7cfef/data/enrichment.benchmark-v2.yml
-upstream_sha: 7a064bf0
+upstream_sha: 4ed7cfef
-imported_at: 2026-08-08
+imported_at: 2026-10-26
 prompt_class: catalogue
 upstream_changes: accepted
 author: upstream
@@ -39,13 +39,55 @@ resources:
    last_checked: 2026-08-08
    metadata_sources:
      - https://lincsproject.org/LINCS/tools/workflows/find-the-best-place-to-obtain-the-lincs-l1000-data
    dataset_profile:
      year: 2017
      readout: l1000
      gene_panel: "978 landmark genes"
      time: true
      view_scope: full-resource
      species: [human]
      sample_type: cell-line
      biological_context: perturbation-screen
      perturbation_type: [chemical, genetic]
      genetic_modes: [mixed]
      paired: false
      pre_post: false
      longitudinal: false
      drug_identity: true
      dose: true
      smiles: partial
      clinical_outcome: false
      n_samples: "large-scale perturbational profiles"
      n_cells: null
      access: open
  prism:
    entities: [cell, drug]
    tasks: [drug-response-prediction]
-    last_checked: 2026-08-08
+    last_checked: 2026-10-02
    metadata_sources:
      - https://depmap.org/portal/prism/
    dataset_profile:
      year: 2020
      readout: viability
      n_compounds: ">4,500"
      n_contexts: ">900 cell lines"
      time: false
      view_scope: full-resource
      species: [human]
      sample_type: cell-line
      biological_context: cancer
      perturbation_type: [chemical]
      paired: true
      pre_post: false
      longitudinal: false
      drug_identity: true
      dose: true
      smiles: partial
      clinical_outcome: false
      n_samples: "PRISM pooled viability screens"
      n_cells: null
      access: open
  pharmgkb:
    entities: [drug, gene, phenotype, variant]
@@ -0,0 +1,125 @@
 ---
 title: "Provenance-backed cancer and pharmacogenomics database enrichment batch."
 task: ""
 lineage_type: import
 upstream_source: https://github.com/inoue0426/awesome-computational-biology/blob/4ed7cfef/data/enrichment.cancer-databases-v1.yml
 upstream_sha: 4ed7cfef
 imported_at: 2026-10-26
 prompt_class: catalogue
 upstream_changes: accepted
 author: upstream
 validated: false
 ---
 # Provenance-backed cancer and pharmacogenomics database enrichment batch.
 resources:
  cancer_cell_line_encyclopedia:
    entities: [cell, gene, disease, drug]
    modalities: [genomics, transcriptomics]
    documentation: https://sites.broadinstitute.org/ccle/
    last_checked: 2026-08-08
    metadata_sources:
      - https://sites.broadinstitute.org/ccle/
  catalogue_of_somatic_mutations_in_cancer_cosmic:
    entities: [disease, gene, variant]
    modalities: [genomics]
    documentation: https://cancer.sanger.ac.uk/cosmic
    last_checked: 2026-08-08
    metadata_sources:
      - https://cancer.sanger.ac.uk/cosmic
  cbioportal:
    entities: [disease, gene, variant]
    modalities: [genomics, clinical]
    github: https://github.com/cBioPortal/cbioportal
    documentation: https://www.cbioportal.org/
    last_checked: 2026-08-08
    metadata_sources:
      - https://www.cbioportal.org/
      - https://github.com/cBioPortal/cbioportal
  dependency_map_depmap:
    entities: [cell, gene, disease, drug]
    modalities: [genomics]
    documentation: https://depmap.org/portal/
    last_checked: 2026-08-08
    metadata_sources:
      - https://depmap.org/portal/
  dgidb:
    entities: [drug, gene]
    modalities: [knowledge-graph]
    documentation: https://www.dgidb.org/
    last_checked: 2026-08-08
    metadata_sources:
      - https://www.dgidb.org/
  drug_mechanism_database_drugmechdb:
    entities: [drug, disease, gene, pathway]
    modalities: [knowledge-graph]
    github: https://github.com/SuLab/DrugMechDB
    last_checked: 2026-08-08
    metadata_sources:
      - https://github.com/SuLab/DrugMechDB
  drug_repurposing_hub:
    entities: [drug, protein, disease]
    modalities: [chemical-structure]
    documentation: https://repo-hub.broadinstitute.org/repurposing
    last_checked: 2026-08-08
    metadata_sources:
      - https://repo-hub.broadinstitute.org/repurposing
  drugcentral:
    entities: [drug, protein, disease]
    modalities: [chemical-structure]
    documentation: https://drugcentral.org/
    last_checked: 2026-08-08
    metadata_sources:
      - https://drugcentral.org/
  drugtargetcommons:
    entities: [drug, protein]
    modalities: [chemical-structure]
    documentation: https://drugtargetcommons.fimm.fi/
    last_checked: 2026-08-08
    metadata_sources:
      - https://drugtargetcommons.fimm.fi/
  comparative_toxicogenomics_database:
    entities: [compound, gene, disease]
    modalities: [knowledge-graph]
    documentation: https://ctdbase.org/
    last_checked: 2026-08-08
    metadata_sources:
      - https://ctdbase.org/
  the_cancer_genome_atlas_tcga:
    entities: [disease, gene]
    modalities: [clinical, genomics]
    last_checked: 2026-10-02
    metadata_sources:
      - https://www.cancer.gov/ccg/research/genome-sequencing/tcga
    dataset_profile:
      year: 2018
      readout: multi-omics
      n_profiles: ">11,000 patients"
      n_contexts: 33
      gene_panel: "transcriptome-wide plus multi-omics"
      time: false
      view_scope: full-resource
      species: [human]
      sample_type: patient
      biological_context: cancer
      perturbation_type: []
      paired: false
      pre_post: false
      longitudinal: false
      drug_identity: limited
      dose: false
      smiles: false
      clinical_outcome: true
      n_samples: ">11,000 patients across 33 cancer types"
      n_cells: null
      access: mixed
@@ -0,0 +1,785 @@
 ---
 title: "Foundation Model Explorer v1 metadata."
 task: ""
 lineage_type: import
 upstream_source: https://github.com/inoue0426/awesome-computational-biology/blob/4ed7cfef/data/enrichment.foundation-explorer-v1.yml
 upstream_sha: 4ed7cfef
 imported_at: 2026-10-26
 prompt_class: unknown
 upstream_changes: accepted
 author: upstream
 validated: false
 ---
 # Foundation Model Explorer v1 metadata.
 # Fine-grained modalities intentionally distinguish assay/data types for navigation.
 resources:
  scfoundation:
    foundation_profile:
      year: 2024
      modalities: [scrna]
      params: "100M"
      params_millions: 100
      pretraining_scale: ">50M human single-cell transcriptomes"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: limited
      spatial: false
  scgpt:
    foundation_profile:
      year: 2024
      modalities: [scrna, multi-omics]
      pretraining_scale: ">33M single-cell profiles"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: true
      spatial: limited
  geneformer:
    foundation_profile:
      year: 2024
      modalities: [scrna]
      params: "10M / 104M / 316M"
      params_millions: 316
      pretraining_scale: "~104M human single-cell transcriptomes in V2; original corpus ~30M"
      species: [human]
      zero_shot: true
      finetunable: true
      weights: true
      code: true
      perturbation: true
      spatial: false
  bulkformer:
    foundation_profile:
      year: 2025
      modalities: [bulk-rna]
      pretraining_scale: "bulk RNA-seq compendium"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  scbert:
    foundation_profile:
      year: 2022
      modalities: [scrna]
      pretraining_scale: "large-scale scRNA-seq corpus"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  cellplm:
    foundation_profile:
      year: 2023
      modalities: [scrna]
      pretraining_scale: "multi-dataset single-cell pretraining"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: limited
      spatial: false
  uce:
    foundation_profile:
      year: 2026
      modalities: [scrna]
      params: ">650M"
      params_millions: 650
      pretraining_scale: ">36M cells across >300 datasets"
      species: [multi-species]
      zero_shot: true
      finetunable: limited
      weights: true
      code: true
      perturbation: false
      spatial: false
  saturn:
    foundation_profile:
      year: 2024
      modalities: [scrna, protein]
      pretraining_scale: "cross-species single-cell atlases with protein-language-model embeddings"
      species: [multi-species]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  cancerfoundation:
    foundation_profile:
      year: 2025
      modalities: [scrna]
      pretraining_scale: "curated malignant-cell single-cell corpus"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: limited
      spatial: false
  cellfm:
    foundation_profile:
      year: 2025
      modalities: [scrna]
      params: "800M"
      params_millions: 800
      pretraining_scale: "100M human cells"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: true
      spatial: false
  scprint:
    foundation_profile:
      year: 2025
      modalities: [scrna]
      pretraining_scale: "large multi-dataset single-cell corpus"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: limited
      spatial: false
  scprint_2:
    foundation_profile:
      year: 2026
      modalities: [scrna]
      pretraining_scale: "350M+ cells; 22K+ datasets; 16 species"
      species: [multi-species]
      zero_shot: true
      finetunable: true
      weights: true
      code: true
      perturbation: limited
      spatial: false
  nicheformer:
    foundation_profile:
      year: 2025
      modalities: [scrna, spatial]
      params: "49.3M"
      params_millions: 49.3
      pretraining_scale: "SpatialCorpus-110M"
      species: [human, mouse]
      zero_shot: true
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: true
  scgpt_spatial:
    foundation_profile:
      year: 2025
      modalities: [scrna, spatial, multi-omics]
      pretraining_scale: "continued pretraining from scGPT with spatial transcriptomics"
      species: [human, mouse]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: true
  scmulan:
    foundation_profile:
      year: 2024
      modalities: [scrna, atac, proteomics, multi-omics]
      pretraining_scale: "~10M cells across transcriptomics, epigenomics, and proteomics"
      species: [human, mouse]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  genecompass:
    foundation_profile:
      year: 2024
      modalities: [scrna, dna, multi-omics]
      pretraining_scale: "120M+ cells across multiple species"
      species: [multi-species]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: limited
      spatial: false
  titan:
    foundation_profile:
      year: 2025
      modalities: [pathology]
      pretraining_scale: "whole-slide pathology with multimodal image-language supervision"
      species: [human]
      zero_shot: true
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  virchow2:
    foundation_profile:
      year: 2024
      modalities: [pathology]
      params: "632M"
      params_millions: 632
      pretraining_scale: "3.1M whole-slide images"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: true
      code: limited
      perturbation: false
      spatial: false
  h_optimus_0:
    foundation_profile:
      year: 2024
      modalities: [pathology]
      params: "1.1B"
      params_millions: 1100
      pretraining_scale: "large multi-center histopathology corpus"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: true
      code: limited
      perturbation: false
      spatial: false
  h_optimus_1:
    foundation_profile:
      year: 2025
      modalities: [pathology]
      params: "1.1B"
      params_millions: 1100
      pretraining_scale: ">1M slides; >800K patients"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: true
      code: limited
      perturbation: false
      spatial: false
  uni2_h:
    foundation_profile:
      year: 2025
      modalities: [pathology]
      params: "~1B"
      params_millions: 1000
      pretraining_scale: "large-scale histopathology pretraining"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: true
      code: limited
      perturbation: false
      spatial: false
  phikon_v2:
    foundation_profile:
      year: 2025
      modalities: [pathology]
      pretraining_scale: "large-scale histopathology pretraining"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: true
      code: limited
      perturbation: false
      spatial: false
  genbio_pathfm:
    foundation_profile:
      year: 2026
      modalities: [pathology]
      params: "1.1B"
      params_millions: 1100
      pretraining_scale: "public histopathology corpus with morphology-aware curation"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  uni_mol:
    foundation_profile:
      year: 2023
      modalities: [chemical, protein]
      pretraining_scale: "large-scale 3D molecular and pocket pretraining"
      species: [multi-species]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  uni_mol2:
    foundation_profile:
      year: 2024
      modalities: [chemical]
      params: "up to 1.1B"
      params_millions: 1100
      pretraining_scale: "800M molecular conformations"
      species: []
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  chemfm:
    foundation_profile:
      year: 2024
      modalities: [chemical]
      params: "1B / 3B"
      params_millions: 3000
      pretraining_scale: "178M molecules"
      species: []
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  esm_cambrian_esm_c:
    foundation_profile:
      year: 2024
      modalities: [protein]
      params: "300M / 600M / 6B"
      params_millions: 6000
      pretraining_scale: "evolutionary-scale protein sequence corpus"
      species: [multi-species]
      zero_shot: true
      finetunable: true
      weights: limited
      code: true
      perturbation: false
      spatial: false
  esm3:
    foundation_profile:
      year: 2024
      modalities: [protein]
      params: "1.4B / 7B / 98B"
      params_millions: 98000
      pretraining_scale: "multimodal protein sequence, structure, and function corpus"
      species: [multi-species]
      zero_shot: true
      finetunable: limited
      weights: limited
      code: true
      perturbation: false
      spatial: false
  boltz_2:
    foundation_profile:
      year: 2025
      modalities: [protein, chemical]
      pretraining_scale: "biomolecular complex structure and affinity data"
      species: [multi-species]
      zero_shot: true
      finetunable: limited
      weights: true
      code: true
      perturbation: false
      spatial: false
  protenix:
    foundation_profile:
      year: 2025
      modalities: [protein, chemical]
      pretraining_scale: "biomolecular structure pretraining"
      species: [multi-species]
      zero_shot: true
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  rna_fm:
    foundation_profile:
      year: 2022
      modalities: [rna]
      params: "99M"
      params_millions: 99
      pretraining_scale: "23.7M non-coding RNA sequences"
      species: [multi-species]
      zero_shot: true
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  rinalmo:
    foundation_profile:
      year: 2025
      modalities: [rna]
      params: "33.5M / 148M / 650M"
      params_millions: 650
      pretraining_scale: "36M non-coding RNA sequences"
      species: [multi-species]
      zero_shot: true
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  evo_2:
    foundation_profile:
      year: 2026
      modalities: [dna]
      params: "1B / 7B / 40B"
      params_millions: 40000
      pretraining_scale: "9T DNA base pairs"
      species: [multi-species]
      zero_shot: true
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  alphagenome:
    foundation_profile:
      year: 2025
      modalities: [dna]
      pretraining_scale: "long-context regulatory genomics corpus"
      species: [human, mouse]
      zero_shot: true
      finetunable: unknown
      weights: no
      code: limited
      perturbation: false
      spatial: false
  moderngena:
    foundation_profile:
      year: 2026
      modalities: [dna]
      pretraining_scale: "hundreds of vertebrate genome assemblies"
      species: [multi-species]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  nucleotide_transformer:
    foundation_profile:
      year: 2025
      modalities: [dna]
      params: "50M – 2.5B"
      params_millions: 2500
      pretraining_scale: "3,202 human genomes and up to 850 species"
      species: [multi-species]
      zero_shot: true
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  dnabert_2:
    foundation_profile:
      year: 2023
      modalities: [dna]
      params: "117M"
      params_millions: 117
      pretraining_scale: "multi-species GenBank genome corpus"
      species: [multi-species]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  caduceus:
    foundation_profile:
      year: 2024
      modalities: [dna]
      pretraining_scale: "long-range genomic sequence pretraining"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  hyenadna:
    foundation_profile:
      year: 2023
      modalities: [dna]
      pretraining_scale: "long-context human genome sequence"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  gpn_genomic_pre_trained_network:
    foundation_profile:
      year: 2023
      modalities: [dna]
      pretraining_scale: "genomic sequence pretraining for variant-effect prediction"
      species: [multi-species]
      zero_shot: true
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  methylgpt:
    foundation_profile:
      year: 2024
      modalities: [methylation]
      params: "3M / 7M / 15M"
      params_millions: 15
      pretraining_scale: "154,063 human methylomes after QC; 5,281 datasets; 49,156 CpG sites"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  cpgpt:
    foundation_profile:
      year: 2025
      modalities: [methylation]
      pretraining_scale: "150K+ methylation samples across >2,000 datasets"
      species: [human]
      zero_shot: true
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  mutationprojector:
    foundation_profile:
      year: 2026
      modalities: [mutation, copy-number]
      pretraining_scale: ">30K tumors with somatic genomic alterations including copy-number changes"
      species: [human]
      zero_shot: limited
      finetunable: unknown
      weights: unknown
      code: unknown
      perturbation: false
      spatial: false
  casanovo_foundation:
    foundation_profile:
      year: 2025
      modalities: [proteomics]
      pretraining_scale: "30M labeled tandem mass spectra"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  tessera:
    foundation_profile:
      year: 2026
      modalities: [mutation, copy-number]
      pretraining_scale: "TCGA pan-cancer somatic SNVs and copy-number segments"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  epiagent:
    foundation_profile:
      year: 2025
      modalities: [atac]
      pretraining_scale: ">5.4M scATAC-seq cells across 37 tissues or cell lines"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: true
      spatial: false
  scarf:
    foundation_profile:
      year: 2025
      modalities: [scrna, atac, multi-omics]
      pretraining_scale: ">2.7M paired and unpaired single cells across tissues and species"
      species: [multi-species]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: limited
      spatial: false
  scdnam_gpt:
    foundation_profile:
      year: 2025
      modalities: [methylation]
      pretraining_scale: ">1M single-cell methylomes from human and mouse tissues"
      species: [human, mouse]
      zero_shot: true
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  epifoundation:
    foundation_profile:
      year: 2025
      modalities: [atac]
      pretraining_scale: "large-scale scATAC-seq pretraining with peak-to-gene alignment"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: limited
      code: limited
      perturbation: false
      spatial: false
  chromfound:
    foundation_profile:
      year: 2025
      modalities: [atac]
      pretraining_scale: "1.97M scATAC-seq cells; 30 tissues; 6 disease contexts"
      species: [human]
      zero_shot: true
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  atacformer:
    foundation_profile:
      year: 2025
      modalities: [atac]
      pretraining_scale: "large atlas of scATAC-seq experiments"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: unknown
      code: unknown
      perturbation: false
      spatial: false
  clm_x:
    foundation_profile:
      year: 2026
      modalities: [scrna, atac, multi-omics]
      pretraining_scale: "million-scale RNA-only, ATAC-only, and paired RNA-ATAC data"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: unknown
      code: unknown
      perturbation: true
      spatial: false
  captain:
    foundation_profile:
      year: 2026
      modalities: [scrna, proteomics, multi-omics]
      pretraining_scale: "co-assayed single-cell RNA and protein data"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: unknown
      code: unknown
      perturbation: false
      spatial: false
  hicfoundation:
    foundation_profile:
      year: 2026
      modalities: [hic]
      pretraining_scale: "hundreds of Hi-C assays; 118M contact-matrix submatrices"
      species: [multi-species]
      zero_shot: limited
      finetunable: true
      weights: unknown
      code: unknown
      perturbation: false
      spatial: false
  omicsfm:
    foundation_profile:
      year: 2026
      modalities: [proteomics, bulk-rna, scrna]
      params: "4.9M"
      params_millions: 4.9
      pretraining_scale: "48,837 proteomics profiles; 680,216 bulk RNA profiles; 4.55M single cells"
      species: [human]
      zero_shot: limited
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: false
  virtues:
    foundation_profile:
      year: 2026
      modalities: [proteomics, spatial]
      pretraining_scale: "31-32 harmonized spatial-proteomics datasets across technologies"
      species: [human]
      zero_shot: true
      finetunable: true
      weights: true
      code: true
      perturbation: false
      spatial: true
@@ -0,0 +1,232 @@
 ---
 title: "Provenance-backed foundation model enrichment batch 3."
 task: ""
 lineage_type: import
 upstream_source: https://github.com/inoue0426/awesome-computational-biology/blob/4ed7cfef/data/enrichment.foundation-models-v3.yml
 upstream_sha: 4ed7cfef
 imported_at: 2026-10-26
 prompt_class: catalogue
 upstream_changes: accepted
 author: upstream
 validated: false
 ---
 # Provenance-backed foundation model enrichment batch 3.
 resources:
  cellfm:
    entities: [cell, gene]
    methods: [self-supervised-learning, transformer]
    modalities: [single-cell-rna-seq, transcriptomics]
    tasks: [cell-type-annotation, foundation-model-pretraining, perturbation-prediction, representation-learning]
    year: 2025
    github: https://github.com/biomed-AI/CellFM
    paper: https://doi.org/10.1038/s41467-025-59926-5
    last_checked: 2026-10-02
    metadata_sources:
      - https://github.com/biomed-AI/CellFM
      - https://doi.org/10.1038/s41467-025-59926-5
  scprint_2:
    entities: [cell, gene]
    methods: [self-supervised-learning, transformer]
    modalities: [single-cell-rna-seq, transcriptomics]
    tasks: [batch-correction, cell-type-annotation, foundation-model-pretraining, gene-regulatory-network-inference, imputation, representation-learning]
    year: 2026
    github: https://github.com/cantinilab/scPRINT-2
    last_checked: 2026-10-02
    metadata_sources:
      - https://github.com/cantinilab/scPRINT-2
  titan:
    entities: [disease, tissue]
    methods: [contrastive-learning, self-supervised-learning, transformer]
    modalities: [histopathology, imaging]
    tasks: [classification, foundation-model-pretraining, representation-learning]
    year: 2025
    github: https://github.com/mahmoodlab/TITAN
    last_checked: 2026-10-02
    metadata_sources:
      - https://github.com/mahmoodlab/TITAN
  virchow2:
    entities: [disease, tissue]
    methods: [self-supervised-learning, transformer]
    modalities: [histopathology, imaging]
    tasks: [classification, foundation-model-pretraining, representation-learning]
    year: 2024
    documentation: https://huggingface.co/paige-ai/Virchow2
    last_checked: 2026-10-02
    metadata_sources:
      - https://huggingface.co/paige-ai/Virchow2
  h_optimus_0:
    entities: [disease, tissue]
    methods: [self-supervised-learning, transformer]
    modalities: [histopathology, imaging]
    tasks: [classification, foundation-model-pretraining, representation-learning]
    year: 2024
    documentation: https://huggingface.co/bioptimus/H-optimus-0
    last_checked: 2026-10-02
    metadata_sources:
      - https://huggingface.co/bioptimus/H-optimus-0
  h_optimus_1:
    entities: [disease, tissue]
    methods: [self-supervised-learning, transformer]
    modalities: [histopathology, imaging]
    tasks: [classification, foundation-model-pretraining, representation-learning]
    year: 2025
    documentation: https://huggingface.co/bioptimus/H-optimus-1
    last_checked: 2026-10-02
    metadata_sources:
      - https://huggingface.co/bioptimus/H-optimus-1
  uni2_h:
    entities: [disease, tissue]
    methods: [self-supervised-learning, transformer]
    modalities: [histopathology, imaging]
    tasks: [classification, foundation-model-pretraining, representation-learning]
    year: 2025
    documentation: https://huggingface.co/MahmoodLab/UNI2-h
    last_checked: 2026-10-02
    metadata_sources:
      - https://huggingface.co/MahmoodLab/UNI2-h
  phikon_v2:
    entities: [disease, tissue]
    methods: [self-supervised-learning, transformer]
    modalities: [histopathology, imaging]
    tasks: [classification, foundation-model-pretraining, representation-learning]
    year: 2025
    documentation: https://huggingface.co/owkin/phikon-v2
    last_checked: 2026-10-02
    metadata_sources:
      - https://huggingface.co/owkin/phikon-v2
  genbio_pathfm:
    entities: [disease, tissue]
    methods: [self-supervised-learning, transformer]
    modalities: [histopathology, imaging]
    tasks: [classification, foundation-model-pretraining, representation-learning]
    year: 2026
    github: https://github.com/genbio-ai/genbio-pathfm
    paper: https://www.biorxiv.org/content/10.64898/2026.03.17.712534v1
    last_checked: 2026-10-02
    metadata_sources:
      - https://github.com/genbio-ai/genbio-pathfm
      - https://www.biorxiv.org/content/10.64898/2026.03.17.712534v1
  uni_mol2:
    entities: [molecule]
    methods: [self-supervised-learning, transformer]
    modalities: [chemical-structure, molecular-structure]
    tasks: [foundation-model-pretraining, representation-learning]
    year: 2024
    github: https://github.com/deepmodeling/Uni-Mol
    last_checked: 2026-10-02
    metadata_sources:
      - https://github.com/deepmodeling/Uni-Mol
  chemfm:
    entities: [molecule]
    methods: [generative-model, language-model, transformer]
    modalities: [chemical-structure]
    tasks: [foundation-model-pretraining, molecular-generation, representation-learning]
    year: 2024
    github: https://github.com/TheLuoFengLab/ChemFM
    last_checked: 2026-10-02
    metadata_sources:
      - https://github.com/TheLuoFengLab/ChemFM
  esm_cambrian_esm_c:
    entities: [protein]
    methods: [language-model, self-supervised-learning, transformer]
    modalities: [protein-sequence]
    tasks: [foundation-model-pretraining, protein-function-prediction, representation-learning]
    year: 2024
    github: https://github.com/Biohub/esm
    last_checked: 2026-10-02
    metadata_sources:
      - https://github.com/Biohub/esm
      - https://www.evolutionaryscale.ai/blog/esm-cambrian
  boltz_2:
    entities: [molecule, protein, protein-complex]
    methods: [diffusion, generative-model, transformer]
    modalities: [molecular-structure, protein-sequence]
    tasks: [drug-target-interaction, structure-prediction, virtual-screening]
    year: 2025
    github: https://github.com/jwohlwend/boltz
    paper: https://doi.org/10.1101/2025.06.14.659707
    last_checked: 2026-10-02
    metadata_sources:
      - https://github.com/jwohlwend/boltz
      - https://doi.org/10.1101/2025.06.14.659707
  protenix:
    entities: [molecule, protein, protein-complex]
    methods: [diffusion, generative-model, transformer]
    modalities: [molecular-structure, protein-sequence]
    tasks: [drug-target-interaction, structure-prediction]
    year: 2025
    github: https://github.com/bytedance/Protenix
    last_checked: 2026-10-02
    metadata_sources:
      - https://github.com/bytedance/Protenix
  rna_fm:
    entities: [gene, transcript]
    methods: [language-model, self-supervised-learning, transformer]
    modalities: [rna-sequence]
    tasks: [foundation-model-pretraining, representation-learning, structure-prediction]
    year: 2022
    github: https://github.com/ml4bio/RNA-FM
    last_checked: 2026-10-02
    metadata_sources:
      - https://github.com/ml4bio/RNA-FM
  rinalmo:
    entities: [gene, transcript]
    methods: [language-model, self-supervised-learning, transformer]
    modalities: [rna-sequence]
    tasks: [foundation-model-pretraining, representation-learning, structure-prediction]
    year: 2024
    github: https://github.com/lbcb-sci/RiNALMo
    last_checked: 2026-10-02
    metadata_sources:
      - https://github.com/lbcb-sci/RiNALMo
  evo_2:
    entities: [gene, genome, organism]
    methods: [generative-model, language-model]
    modalities: [dna-sequence, genomics]
    tasks: [foundation-model-pretraining, representation-learning]
    year: 2026
    github: https://github.com/arcinstitute/evo2
    paper: https://doi.org/10.1038/s41586-026-10176-5
    last_checked: 2026-10-02
    metadata_sources:
      - https://github.com/arcinstitute/evo2
      - https://doi.org/10.1038/s41586-026-10176-5
  alphagenome:
    entities: [gene, genome, regulatory-element, variant]
    methods: [supervised-learning, transformer]
    modalities: [dna-sequence, epigenomics, genomics]
    tasks: [regression, representation-learning]
    year: 2025
    github: https://github.com/google-deepmind/alphagenome
    last_checked: 2026-10-02
    metadata_sources:
      - https://github.com/google-deepmind/alphagenome
  moderngena:
    entities: [gene, genome, regulatory-element]
    methods: [language-model, self-supervised-learning, transformer]
    modalities: [dna-sequence, genomics]
    tasks: [classification, foundation-model-pretraining, regression, representation-learning]
    year: 2026
    github: https://github.com/AIRI-Institute/GENA_LM
    last_checked: 2026-10-02
    metadata_sources:
      - https://github.com/AIRI-Institute/GENA_LM
@@ -0,0 +1,292 @@
 ---
 title: "Genetic perturbation Dataset Explorer metadata; artifact sync verified."
 task: ""
 lineage_type: import
 upstream_source: https://github.com/inoue0426/awesome-computational-biology/blob/4ed7cfef/data/enrichment.genetic-perturbation-v1.yml
 upstream_sha: 4ed7cfef
 imported_at: 2026-10-26
 prompt_class: catalogue
 upstream_changes: accepted
 author: upstream
 validated: false
 ---
 # Genetic perturbation Dataset Explorer metadata; artifact sync verified.
 # Curated genetic perturbation datasets for Dataset Explorer.
 resources:
  dixit_et_al_2016_perturb_seq:
    paper: https://doi.org/10.1016/j.cell.2016.11.038
    last_checked: 2026-10-02
    metadata_sources:
      - https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE90063
      - https://doi.org/10.1016/j.cell.2016.11.038
    dataset_profile:
      year: 2016
      view_scope: full-study
      species: [human, mouse]
      sample_type: cell-line+primary
      biological_context: immune-and-cell-line
      perturbation_type: [genetic]
      genetic_modes: [crispr-ko, combinatorial]
      readout: single-cell-rna
      n_profiles: "~200,000 cells"
      n_perturbations: "pooled CRISPR-KO across six screens"
      n_contexts: 6
      gene_panel: transcriptome-wide
      paired: true
      pre_post: false
      longitudinal: false
      drug_identity: false
      dose: false
      time: true
      smiles: false
      clinical_outcome: false
      n_samples: "six pooled screens across K562 and dendritic-cell conditions"
      n_cells: "~200,000"
      access: open
  adamson_et_al_2016_perturb_seq:
    paper: https://doi.org/10.1016/j.cell.2016.11.048
    last_checked: 2026-10-02
    metadata_sources:
      - https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE90546
    dataset_profile:
      year: 2016
      view_scope: full-study
      species: [human]
      sample_type: cell-line
      biological_context: unfolded-protein-response
      perturbation_type: [genetic]
      genetic_modes: [crispri, combinatorial]
      readout: single-cell-rna
      n_profiles: "~86,000 cells"
      n_perturbations: "~100 UPR hits plus combinatorial CRISPRi perturbations"
      n_contexts: 3
      gene_panel: transcriptome-wide
      paired: true
      pre_post: false
      longitudinal: false
      drug_identity: false
      dose: false
      time: false
      smiles: false
      clinical_outcome: false
      n_samples: "three pooled CRISPRi Perturb-seq experiments in K562"
      n_cells: "~86,000"
      access: open
  datlinger_et_al_2017_crop_seq:
    paper: https://doi.org/10.1038/nmeth.4177
    last_checked: 2026-10-02
    metadata_sources:
      - https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE92872
      - https://doi.org/10.1038/nmeth.4177
    dataset_profile:
      year: 2017
      view_scope: full-study
      species: [human]
      sample_type: cell-line
      biological_context: t-cell-receptor-signaling
      perturbation_type: [genetic]
      genetic_modes: [crispr-ko]
      readout: single-cell-rna
      n_profiles: 5905
      n_perturbations: "TCR regulators and transcription factors"
      n_contexts: 2
      gene_panel: transcriptome-wide
      paired: true
      pre_post: false
      longitudinal: false
      drug_identity: false
      dose: false
      time: false
      smiles: false
      clinical_outcome: false
      n_samples: "stimulated and unstimulated Jurkat CROP-seq"
      n_cells: 5905
      access: open
  norman_et_al_2019_crispra_perturb_seq:
    paper: https://doi.org/10.1126/science.aax4438
    last_checked: 2026-10-02
    metadata_sources:
      - https://figshare.com/articles/dataset/Norman_et_al_2019_Perturb-seq/27766323
      - https://doi.org/10.1126/science.aax4438
    dataset_profile:
      year: 2019
      view_scope: full-study
      species: [human]
      sample_type: cell-line
      biological_context: genetic-interactions
      perturbation_type: [genetic]
      genetic_modes: [crispra, combinatorial]
      readout: single-cell-rna
      n_profiles: "~110,000 cells"
      n_perturbations: 287
      n_contexts: 1
      gene_panel: transcriptome-wide
      paired: true
      pre_post: false
      longitudinal: false
      drug_identity: false
      dose: false
      time: false
      smiles: false
      clinical_outcome: false
      n_samples: "single-gene and pairwise CRISPRa perturbations in K562"
      n_cells: "~110,000"
      access: open
  gasperini_et_al_2019_crispri_enhancer_screen:
    paper: https://doi.org/10.1016/j.cell.2018.11.029
    last_checked: 2026-10-02
    metadata_sources:
      - https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE120861
    dataset_profile:
      year: 2019
      view_scope: full-study
      species: [human]
      sample_type: cell-line
      biological_context: enhancer-gene-regulation
      perturbation_type: [genetic]
      genetic_modes: [crispri, enhancer-targeting]
      readout: single-cell-rna
      n_profiles: 254974
      n_perturbations: 5920
      n_contexts: 3
      gene_panel: transcriptome-wide
      paired: true
      pre_post: false
      longitudinal: false
      drug_identity: false
      dose: false
      time: false
      smiles: false
      clinical_outcome: false
      n_samples: "three CRISPRi screens in K562, including an at-scale enhancer screen"
      n_cells: 254974
      access: open
  replogle_et_al_2022_k562_genome_wide_perturb_seq:
    paper: https://doi.org/10.1016/j.cell.2022.05.013
    last_checked: 2026-10-02
    metadata_sources:
      - https://plus.figshare.com/articles/dataset/_Mapping_information-rich_genotype-phenotype_landscapes_with_genome-scale_Perturb-seq_Replogle_et_al_2022_processed_Perturb-seq_datasets/20029387
      - https://gwps.wi.mit.edu/
    dataset_profile:
      year: 2022
      view_scope: full-study
      species: [human]
      sample_type: cell-line
      biological_context: genome-wide-functional-genomics
      perturbation_type: [genetic]
      genetic_modes: [crispri]
      readout: single-cell-rna
      n_profiles: "genome-scale; mean ~183 cells per detected perturbation"
      n_perturbations: "9,866 target genes; 11,258 detected sgRNA-pair perturbations"
      n_contexts: 1
      gene_panel: transcriptome-wide
      paired: true
      pre_post: false
      longitudinal: false
      drug_identity: false
      dose: false
      time: true
      smiles: false
      clinical_outcome: false
      n_samples: "K562 CRISPRi genome-wide screen sampled at day 8"
      n_cells: "mean ~183 cells per detected perturbation"
      access: open
  replogle_et_al_2022_k562_essential_perturb_seq:
    paper: https://doi.org/10.1016/j.cell.2022.05.013
    last_checked: 2026-10-02
    metadata_sources:
      - https://plus.figshare.com/articles/dataset/_Mapping_information-rich_genotype-phenotype_landscapes_with_genome-scale_Perturb-seq_Replogle_et_al_2022_processed_Perturb-seq_datasets/20029387
      - https://gwps.wi.mit.edu/
    dataset_profile:
      year: 2022
      view_scope: full-study
      species: [human]
      sample_type: cell-line
      biological_context: essential-gene-screen
      perturbation_type: [genetic]
      genetic_modes: [crispri]
      readout: single-cell-rna
      n_profiles: 310385
      n_perturbations: 2057
      n_contexts: 1
      gene_panel: transcriptome-wide
      paired: true
      pre_post: false
      longitudinal: false
      drug_identity: false
      dose: false
      time: true
      smiles: false
      clinical_outcome: false
      n_samples: "K562 CRISPRi essential-gene screen sampled at day 6"
      n_cells: 310385
      access: open
  replogle_et_al_2022_rpe1_essential_perturb_seq:
    paper: https://doi.org/10.1016/j.cell.2022.05.013
    last_checked: 2026-10-02
    metadata_sources:
      - https://plus.figshare.com/articles/dataset/_Mapping_information-rich_genotype-phenotype_landscapes_with_genome-scale_Perturb-seq_Replogle_et_al_2022_processed_Perturb-seq_datasets/20029387
      - https://gwps.wi.mit.edu/
    dataset_profile:
      year: 2022
      view_scope: full-study
      species: [human]
      sample_type: cell-line
      biological_context: essential-gene-screen
      perturbation_type: [genetic]
      genetic_modes: [crispri]
      readout: single-cell-rna
      n_profiles: 247914
      n_perturbations: 2393
      n_contexts: 1
      gene_panel: transcriptome-wide
      paired: true
      pre_post: false
      longitudinal: false
      drug_identity: false
      dose: false
      time: true
      smiles: false
      clinical_outcome: false
      n_samples: "RPE1 CRISPRi essential-gene screen sampled at day 7"
      n_cells: 247914
      access: open
  frangieh_et_al_2021_perturb_cite_seq:
    paper: https://doi.org/10.1038/s41588-021-00779-1
    last_checked: 2026-10-02
    metadata_sources:
      - https://singlecell.broadinstitute.org/single_cell/study/SCP1064/multi-modal-pooled-perturb-cite-seq-screens-in-patient-models-define-novel-mechanisms-of-cancer-immune-evasion
      - https://doi.org/10.1038/s41588-021-00779-1
    dataset_profile:
      year: 2021
      view_scope: full-study
      species: [human]
      sample_type: patient-derived-cell-line
      biological_context: melanoma-immunotherapy
      perturbation_type: [genetic]
      genetic_modes: [crispr-ko]
      readout: cite-seq
      n_profiles: ">200,000 cells"
      n_perturbations: 248
      n_contexts: 3
      gene_panel: "transcriptome + surface proteins"
      paired: true
      pre_post: false
      longitudinal: false
      drug_identity: false
      dose: false
      time: false
      smiles: false
      clinical_outcome: false
      n_samples: "patient-derived melanoma under control, IFN-gamma, and TIL co-culture conditions"
      n_cells: ">200,000"
      access: mixed
@@ -0,0 +1,285 @@
 ---
 title: "Method Explorer v1; changes to this file trigger artifact regeneration."
 task: ""
 lineage_type: import
 upstream_source: https://github.com/inoue0426/awesome-computational-biology/blob/4ed7cfef/data/enrichment.method-explorer-v1.yml
 upstream_sha: 4ed7cfef
 imported_at: 2026-10-26
 prompt_class: unknown
 upstream_changes: accepted
 author: upstream
 validated: false
 ---
 # Method Explorer v1; changes to this file trigger artifact regeneration.
 # Structured comparison metadata for Method Explorer v1.
 resources:
  bison:
    method_profile:
      task_family: drug-perturbation
      year: 2026
      input: bulk/pseudobulk transcriptomics
      drug_representation: Morgan fingerprints + chemical tokens
      context_representation: control-derived context features + embeddings
      output: transcriptomic response
      unseen_drug: yes
      unseen_context: limited
      dose: true
      time: true
      patient_transfer: false
      code: false
  txpert:
    method_profile:
      task_family: drug-perturbation
      year: 2026
      input: transcriptomics + biological knowledge
      drug_representation: knowledge-graph-informed perturbation representation
      context_representation: cellular context representation
      output: transcriptomic response
      unseen_drug: yes
      unseen_context: yes
      dose: false
      time: false
      patient_transfer: false
      code: true
  lpm:
    method_profile:
      task_family: drug-perturbation
      year: 2025
      input: heterogeneous perturbation experiments
      drug_representation: learned perturbation representation
      context_representation: learned context representation
      output: molecular state
      unseen_drug: yes
      unseen_context: yes
      dose: limited
      time: limited
      patient_transfer: false
      code: true
  state:
    method_profile:
      task_family: drug-perturbation
      year: 2026
      input: sets of cellular expression profiles
      drug_representation: perturbation embedding
      context_representation: cellular state
      output: transcriptomic response
      unseen_drug: limited
      unseen_context: yes
      dose: limited
      time: limited
      patient_transfer: false
      code: true
  chemcpa:
    method_profile:
      task_family: drug-perturbation
      year: 2022
      input: single-cell expression + covariates
      drug_representation: RDKit2D descriptors
      context_representation: cell covariates + basal state
      output: transcriptomic response
      unseen_drug: yes
      unseen_context: limited
      dose: true
      time: limited
      patient_transfer: false
      code: true
  cmonge:
    method_profile:
      task_family: drug-perturbation
      year: 2026
      input: source/target expression distributions
      drug_representation: RDKit2D descriptors
      context_representation: matched control distribution
      output: transported expression distribution
      unseen_drug: yes
      unseen_context: yes
      dose: true
      time: limited
      patient_transfer: false
      code: true
  prnet:
    method_profile:
      task_family: drug-perturbation
      year: 2024
      input: basal expression + perturbation
      drug_representation: FCFP4 fingerprint
      context_representation: matched basal expression
      output: transcriptomic response
      unseen_drug: yes
      unseen_context: limited
      dose: true
      time: limited
      patient_transfer: false
      code: true
  biolord:
    method_profile:
      task_family: drug-perturbation
      year: 2024
      input: expression + known attributes
      drug_representation: chemical descriptors
      context_representation: unknown latent attributes
      output: transcriptomic response
      unseen_drug: yes
      unseen_context: yes
      dose: true
      time: true
      patient_transfer: false
      code: true
  perturbnet:
    method_profile:
      task_family: drug-perturbation
      year: 2025
      input: single-cell expression distributions
      drug_representation: molecular representation
      context_representation: conditional latent state
      output: cell-state distribution
      unseen_drug: yes
      unseen_context: limited
      dose: limited
      time: limited
      patient_transfer: false
      code: true
  cellflow:
    method_profile:
      task_family: drug-perturbation
      year: 2025
      input: control/treated expression distributions
      drug_representation: molecular features
      context_representation: matched control distribution
      output: transported cell-state distribution
      unseen_drug: yes
      unseen_context: yes
      dose: true
      time: true
      patient_transfer: false
      code: true
  prophet:
    method_profile:
      task_family: drug-perturbation
      year: 2024
      input: control expression + perturbation
      drug_representation: learned chemical representation
      context_representation: control-derived context
      output: gene-expression response
      unseen_drug: yes
      unseen_context: yes
      dose: true
      time: true
      patient_transfer: false
      code: true
  prepr_ct:
    method_profile:
      task_family: drug-perturbation
      year: 2026
      input: aggregate expression + context graph
      drug_representation: RDKit descriptors
      context_representation: context-specific gene graph
      output: transcriptomic response
      unseen_drug: yes
      unseen_context: limited
      dose: true
      time: true
      patient_transfer: false
      code: true
  xpert:
    method_profile:
      task_family: drug-perturbation
      year: 2026
      input: transcriptomics + heterogeneous graph
      drug_representation: Uni-Mol + knowledge graph
      context_representation: basal expression + gene embeddings
      output: transcriptomic response
      unseen_drug: yes
      unseen_context: yes
      dose: true
      time: true
      patient_transfer: false
      code: true
  paso:
    method_profile:
      task_family: drug-response-prediction
      year: 2025
      input: pathway-aware multi-omics
      drug_representation: SMILES-derived representation
      context_representation: pathway-difference representation
      output: cell-line drug response
      unseen_drug: limited
      unseen_context: limited
      dose: false
      time: false
      patient_transfer: false
      code: true
  dtlcdr:
    method_profile:
      task_family: drug-response-prediction
      year: 2025
      input: multimodal tumor/cell-line features + targets
      drug_representation: target-aware drug representation
      context_representation: biological sample representation
      output: preclinical/clinical response
      unseen_drug: yes
      unseen_context: limited
      dose: false
      time: false
      patient_transfer: true
      code: false
  therapi:
    method_profile:
      task_family: drug-response-prediction
      year: 2026
      input: tumor transcriptomics + cell-line knowledge
      drug_representation: perturbation/gene-level representation
      context_representation: aligned tumor/cell-line representation
      output: patient drug response
      unseen_drug: limited
      unseen_context: yes
      dose: false
      time: false
      patient_transfer: true
      code: true
  expresso:
    method_profile:
      task_family: drug-response-prediction
      year: 2026
      input: pretreatment tumor transcriptomics
      drug_representation: drug targets + biomarkers
      context_representation: tumor transcriptomic context
      output: patient treatment response
      unseen_drug: limited
      unseen_context: yes
      dose: false
      time: false
      patient_transfer: true
      code: false
  perturbrx:
    method_profile:
      task_family: drug-response-prediction
      year: 2026
      input: pretreatment tumor RNA-seq + perturbation pretraining
      drug_representation: drug embedding + treatment-conditioned transition
      context_representation: patient latent state
      output: patient treatment response
      unseen_drug: limited
      unseen_context: yes
      dose: true
      time: false
      patient_transfer: true
      code: false
@@ -0,0 +1,321 @@
 ---
 title: "Includes Bison benchmark coverage metadata."
 task: ""
 lineage_type: import
 upstream_source: https://github.com/inoue0426/awesome-computational-biology/blob/4ed7cfef/data/enrichment.perturbation-datasets-v1.yml
 upstream_sha: 4ed7cfef
 imported_at: 2026-10-26
 prompt_class: catalogue
 upstream_changes: accepted
 author: upstream
 validated: false
 ---
 # Includes Bison benchmark coverage metadata.
 # Curated perturbation, patient, and PDX dataset profiles for Dataset Explorer.
 resources:
  tahoe_100m:
    dataset_profile:
      year: 2025
      view_scope: bison-benchmark-view
      readout: single-cell-rna
      n_profiles: 67018
      n_compounds: 376
      n_contexts: 50
      gene_panel: "20,573 full / 2,000 HVG (Bison benchmark view)"
      time: true
      species: [human]
      sample_type: cell-line
      biological_context: cancer
      perturbation_type: [chemical]
      paired: false
      pre_post: false
      longitudinal: false
      drug_identity: true
      dose: true
      smiles: partial
      clinical_outcome: false
      n_samples: ">100 million profiles; 50 cancer cell lines; ~1,100 compounds"
      n_cells: ">100 million"
      access: open
    last_checked: 2026-10-02
    metadata_sources:
      - https://huggingface.co/datasets/tahoebio/Tahoe-100M
  sci_plex:
    dataset_profile:
      year: 2020
      view_scope: bison-benchmark-view
      readout: pseudobulk-rna
      n_profiles: 4974
      n_compounds: 186
      n_contexts: 3
      gene_panel: "20,251 full / 2,000 HVG (Bison benchmark view)"
      time: true
      species: [human, mouse]
      sample_type: cell-line
      biological_context: perturbation-screen
      perturbation_type: [chemical]
      paired: false
      pre_post: false
      longitudinal: false
      drug_identity: true
      dose: true
      smiles: partial
      clinical_outcome: false
      n_samples: "~5,000 conditions; 188 compounds across 3 human cancer cell lines"
      n_cells: 649340
      access: open
    last_checked: 2026-10-02
    metadata_sources:
      - https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE139944
      - https://pubmed.ncbi.nlm.nih.gov/31806696/
  mix_seq:
    dataset_profile:
      year: 2020
      view_scope: full-resource
      readout: single-cell-rna
      n_compounds: 13
      n_contexts: "100+ cell lines"
      gene_panel: "transcriptome-wide"
      time: true
      species: [human]
      sample_type: cell-line
      biological_context: cancer
      perturbation_type: [chemical, genetic]
      genetic_modes: [mixed]
      paired: false
      pre_post: false
      longitudinal: true
      drug_identity: true
      dose: true
      smiles: partial
      clinical_outcome: false
      n_samples: "pools of up to 100+ cancer cell lines; 13-drug demonstration"
      n_cells: "single-cell RNA-seq"
      access: open
    last_checked: 2026-10-02
    metadata_sources:
      - https://www.nature.com/articles/s41467-020-17440-w
  pdx_atlas:
    dataset_profile:
      year: 2026
      view_scope: current-portal
      readout: multi-omics
      n_profiles: ">50 PDXs"
      n_contexts: "30 patients"
      gene_panel: "multi-omics"
      time: false
      species: [human, mouse]
      sample_type: pdx
      biological_context: cancer
      perturbation_type: [chemical]
      paired: true
      pre_post: false
      longitudinal: false
      drug_identity: true
      dose: false
      smiles: partial
      clinical_outcome: false
      n_samples: ">50 PDXs from 30 prostate cancer patients in the current portal"
      n_cells: null
      access: open
    last_checked: 2026-10-02
    metadata_sources:
      - https://muralportal.pdxatlas.org/
  gse191127_breast_cancer_pre_post_chemotherapy:
    dataset_profile:
      year: 2022
      view_scope: study-cohort
      readout: bulk-rna
      n_profiles: "317 pretreatment; 22 matched pre/post deeply profiled"
      n_contexts: 1
      gene_panel: "transcriptome-wide"
      time: true
      species: [human]
      sample_type: patient
      biological_context: breast-cancer
      perturbation_type: [chemical]
      paired: true
      pre_post: true
      longitudinal: true
      drug_identity: limited
      dose: false
      smiles: partial
      clinical_outcome: true
      n_samples: "317 pretreatment biopsies; 22 matched pre/post tumors with deep profiling"
      n_cells: null
      access: open
    last_checked: 2026-10-02
    metadata_sources:
      - https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE191127
  op3:
    dataset_profile:
      year: 2024
      view_scope: bison-benchmark-view
      readout: bulk-rna
      n_profiles: 1813
      n_compounds: 138
      n_contexts: 4
      gene_panel: "5,198 full / 2,000 HVG"
      time: true
      species: [human]
      sample_type: cell-line
      biological_context: perturbation-screen
      perturbation_type: [chemical]
      paired: false
      pre_post: false
      longitudinal: false
      drug_identity: true
      dose: true
      smiles: true
      clinical_outcome: false
      n_samples: "1,813 profiles; 138 molecules; 4 contexts"
      n_cells: null
      access: open
    last_checked: 2026-10-02
    metadata_sources:
      - https://github.com/theislab/Chem-PerturBridge
  vcpi_0001:
    dataset_profile:
      view_scope: bison-benchmark-view
      readout: bulk-rna
      n_profiles: 27517
      n_compounds: 2272
      n_contexts: 1
      gene_panel: "20,629 full / 2,000 HVG"
      time: true
      species: [human]
      sample_type: cell-line
      biological_context: perturbation-screen
      perturbation_type: [chemical]
      paired: false
      pre_post: false
      longitudinal: false
      drug_identity: true
      dose: true
      smiles: true
      clinical_outcome: false
      n_samples: "27,517 profiles; 2,272 molecules; 1 context"
      n_cells: null
      access: open
    last_checked: 2026-10-02
    metadata_sources:
      - https://github.com/theislab/Chem-PerturBridge
  vcpi_0002:
    dataset_profile:
      view_scope: bison-benchmark-view
      readout: bulk-rna
      n_profiles: 18139
      n_compounds: 1488
      n_contexts: 1
      gene_panel: "20,629 full / 2,000 HVG"
      time: true
      species: [human]
      sample_type: cell-line
      biological_context: perturbation-screen
      perturbation_type: [chemical]
      paired: false
      pre_post: false
      longitudinal: false
      drug_identity: true
      dose: true
      smiles: true
      clinical_outcome: false
      n_samples: "18,139 profiles; 1,488 molecules; 1 context"
      n_cells: null
      access: open
    last_checked: 2026-10-02
    metadata_sources:
      - https://github.com/theislab/Chem-PerturBridge
  novartis_perturbation_dataset:
    dataset_profile:
      year: 2025
      view_scope: bison-benchmark-view
      readout: bulk-rna
      n_profiles: 46748
      n_compounds: 3770
      n_contexts: 1
      gene_panel: "20,038 full / 2,000 HVG"
      time: true
      species: [human]
      sample_type: cell-line
      biological_context: perturbation-screen
      perturbation_type: [chemical]
      paired: false
      pre_post: false
      longitudinal: false
      drug_identity: true
      dose: true
      smiles: true
      clinical_outcome: false
      n_samples: "46,748 profiles; 3,770 molecules; 1 context"
      n_cells: null
      access: open
    last_checked: 2026-10-02
    metadata_sources:
      - https://github.com/theislab/Chem-PerturBridge
  lincs_l1000_phase_1:
    dataset_profile:
      year: 2017
      view_scope: bison-benchmark-view
      readout: l1000
      n_profiles: 692787
      n_compounds: 9233
      n_contexts: 70
      gene_panel: "978 landmark genes"
      time: true
      species: [human]
      sample_type: cell-line
      biological_context: perturbation-screen
      perturbation_type: [chemical]
      paired: false
      pre_post: false
      longitudinal: false
      drug_identity: true
      dose: true
      smiles: true
      clinical_outcome: false
      n_samples: "692,787 profiles; 9,233 molecules; 70 contexts; 978 landmark genes"
      n_cells: null
      access: open
    last_checked: 2026-10-02
    metadata_sources:
      - https://github.com/theislab/Chem-PerturBridge
  lincs_l1000_phase_2:
    dataset_profile:
      year: 2017
      view_scope: bison-benchmark-view
      readout: l1000
      n_profiles: 333263
      n_compounds: 1760
      n_contexts: 30
      gene_panel: "978 landmark genes"
      time: true
      species: [human]
      sample_type: cell-line
      biological_context: perturbation-screen
      perturbation_type: [chemical]
      paired: false
      pre_post: false
      longitudinal: false
      drug_identity: true
      dose: true
      smiles: true
      clinical_outcome: false
      n_samples: "333,263 profiles; 1,760 molecules; 30 contexts; 978 landmark genes"
      n_cells: null
      access: open
    last_checked: 2026-10-02
    metadata_sources:
      - https://github.com/theislab/Chem-PerturBridge
@@ -2,9 +2,9 @@
 title: "Provenance-backed drug-response and pharmacogenomics enrichment batch."
 task: ""
 lineage_type: import
-upstream_source: https://github.com/inoue0426/awesome-computational-biology/blob/7a064bf0/data/enrichment.pharmacogenomics-v1.yml
+upstream_source: https://github.com/inoue0426/awesome-computational-biology/blob/4ed7cfef/data/enrichment.pharmacogenomics-v1.yml
-upstream_sha: 7a064bf0
+upstream_sha: 4ed7cfef
-imported_at: 2026-08-08
+imported_at: 2026-10-26
 prompt_class: catalogue
 upstream_changes: accepted
 author: upstream
@@ -20,6 +20,25 @@ resources:
    last_checked: 2026-08-08
    metadata_sources:
      - https://biodev.github.io/BeatAML2/
    dataset_profile:
      year: 2018
      readout: ex-vivo-viability
      time: false
      view_scope: full-resource
      species: [human]
      sample_type: patient
      biological_context: cancer
      perturbation_type: [chemical]
      paired: true
      pre_post: false
      longitudinal: false
      drug_identity: true
      dose: true
      smiles: partial
      clinical_outcome: true
      n_samples: "primary AML patient samples"
      n_cells: null
      access: open
  cancer_therapeutics_response_portal_ctrp:
    entities: [cell, drug]
@@ -27,6 +46,27 @@ resources:
    last_checked: 2026-08-08
    metadata_sources:
      - https://portals.broadinstitute.org/ctrp/
    dataset_profile:
      year: 2013
      readout: viability
      n_compounds: ">400"
      n_contexts: "~900 cell lines"
      time: false
      view_scope: full-resource
      species: [human]
      sample_type: cell-line
      biological_context: cancer
      perturbation_type: [chemical]
      paired: true
      pre_post: false
      longitudinal: false
      drug_identity: true
      dose: true
      smiles: partial
      clinical_outcome: false
      n_samples: "~900 cancer cell lines"
      n_cells: null
      access: open
  bindingdb_curated_sets:
    entities: [molecule, protein]
@@ -71,9 +111,30 @@ resources:
    modalities: [single-cell-rna-seq]
    tasks: [perturbation-prediction]
    github: https://github.com/sanderlab/scPerturb
-    last_checked: 2026-08-08
+    last_checked: 2026-10-02
    metadata_sources:
      - https://github.com/sanderlab/scPerturb
    dataset_profile:
      year: 2024
      readout: single-cell-rna
      gene_panel: "varies by source study"
      time: true
      view_scope: harmonized-collection
      species: [human, mouse]
      sample_type: cell-line
      biological_context: perturbation-collection
      perturbation_type: [chemical, genetic]
      genetic_modes: [mixed]
      paired: false
      pre_post: false
      longitudinal: false
      drug_identity: true
      dose: true
      smiles: partial
      clinical_outcome: false
      n_samples: "scPerturb integrated perturbation datasets"
      n_cells: "multi-study single-cell compendium"
      access: open
  genomics_of_drug_sensitivity_in_cancer_gdsc:
    entities: [cell, drug, gene]
@@ -82,6 +143,26 @@ resources:
    last_checked: 2026-08-08
    metadata_sources:
      - https://www.cancerrxgene.org/
    dataset_profile:
      year: 2012
      readout: viability
      n_contexts: "~1,000 cell lines"
      time: false
      view_scope: full-resource
      species: [human]
      sample_type: cell-line
      biological_context: cancer
      perturbation_type: [chemical]
      paired: true
      pre_post: false
      longitudinal: false
      drug_identity: true
      dose: true
      smiles: partial
      clinical_outcome: false
      n_samples: "~1,000 cancer cell lines"
      n_cells: null
      access: open
  cellminer_cross_database_cellminercdb:
    entities: [cell, drug, gene]
@@ -2,9 +2,9 @@
 title: "Resource Data Schema (`docs/data/resources.json`)"
 task: ""
 lineage_type: import
-upstream_source: https://github.com/inoue0426/awesome-computational-biology/blob/12d87583/docs/data/SCHEMA.md
+upstream_source: https://github.com/inoue0426/awesome-computational-biology/blob/4ed7cfef/docs/data/SCHEMA.md
-upstream_sha: 12d87583
+upstream_sha: 4ed7cfef
-imported_at: 2026-06-26
+imported_at: 2026-10-26
 prompt_class: catalogue
 upstream_changes: accepted
 author: upstream
@@ -52,6 +52,10 @@ When contributing new resources, update `README.md` first, then regenerate artif
 | `api` | boolean | Whether programmatic API access is available. Defaults to `false`. |
 | `paper` | string | DOI or URL to preprint/peer-reviewed publication. |
 | `updated` | string | Last-known update date, recommended `YYYY-MM-DD`. |
 | `dataset_profile` | object | Optional Dataset Explorer metadata for structured comparison of datasets/benchmarks. |
 | `method_profile` | object | Optional Method Explorer metadata for structured comparison of response and perturbation methods. |
 | `foundation_profile` | object | Optional Foundation Model Explorer metadata for modality, model scale, size, and availability. |
 | `agent_profile` | object | Optional Agent Explorer metadata for agent architecture, capabilities, and scientific workflow support. |
 ## Naming and consistency guidance
@@ -79,3 +83,101 @@ When contributing new resources, update `README.md` first, then regenerate artif
  "updated": "2026-01-15"
 }
 ```
 ## Dataset Explorer profile
 Resources that participate in the Dataset Explorer may define a `dataset_profile` object in a modular `data/enrichment.*.yml` file. This keeps README-derived identity fields separate from deeper dataset metadata.
 Supported v1 fields:
 | Field | Type | Meaning |
 |---|---|---|
 | `year` | integer | Representative primary publication or public release year. |
 | `species` | array of strings | Species represented in the dataset. |
 | `sample_type` | string | Primary unit, e.g. `patient` or `cell-line`. |
 | `biological_context` | string | High-level context such as `cancer` or `perturbation-screen`. |
 | `perturbation_type` | array of strings | Chemical, genetic, or other interventions. |
 | `genetic_modes` | array of strings | Genetic perturbation mechanisms, e.g. `crispri`, `crispra`, `crispr-ko`, `enhancer-targeting`, `combinatorial`, or `mixed`. |
 | `paired` | boolean | Whether molecular/sample identity can be linked to the measured response or perturbation condition. |
 | `pre_post` | boolean | Whether matched pre-treatment and post-treatment samples are available. |
 | `longitudinal` | boolean | Whether repeated measurements over time are available as a core design feature. |
 | `drug_identity` | boolean or string | Whether treatment identity is available; strings such as `limited` may encode partial availability. |
 | `dose` | boolean | Whether dose/concentration metadata are available. |
 | `smiles` | boolean or `partial`/`limited` | Whether compound structures are directly available or can only be partially mapped. |
 | `clinical_outcome` | boolean | Whether patient-level clinical outcomes are available. |
 | `n_samples` | string | Human-readable scale summary. |
 | `n_cells` | integer, string, or null | Cell count for single-cell datasets when meaningful. |
 | `n_profiles` | integer, string, or null | Number of profiles in the curated comparison view. |
 | `n_perturbations` | integer, string, or null | Number of perturbation targets/conditions, especially useful for genetic screens. |
 | `n_compounds` | integer, string, or null | Number of compounds/molecules in the curated comparison view. |
 | `n_contexts` | integer, string, or null | Number of cellular or experimental contexts. |
 | `readout` | string | Primary readout, e.g. `bulk-rna`, `pseudobulk-rna`, `single-cell-rna`, or `viability`. |
 | `gene_panel` | string | Gene/readout panel summary, e.g. `20,251 full / 2,000 HVG` or `978 landmark genes`. |
 | `time` | boolean | Whether treatment time metadata are available. |
 | `view_scope` | string | Scope of the displayed scale fields, e.g. `full-resource`, `study-cohort`, or `bison-benchmark-view`. |
 | `access` | string | Dataset-level access summary. |
 The v1 profiles are colocated with their existing modular `data/enrichment.*.yml` owners and rendered at `docs/dataset-explorer.html`.
 ## Method Explorer profile
 Methods shown in the Method Explorer may define a `method_profile` object in a modular enrichment fragment.
 | Field | Type | Meaning |
 |---|---|---|
 | `task_family` | string | High-level task, currently `drug-response-prediction` or `drug-perturbation`. |
 | `year` | integer | Publication/release year used for navigation. |
 | `input` | string | Main biological/model input. |
 | `drug_representation` | string | Chemical or treatment representation. |
 | `context_representation` | string | Cellular, patient, or experimental context representation. |
 | `output` | string | Main prediction target. |
 | `unseen_drug` | boolean/string | Support for unseen-drug generalization; `limited` denotes partial or task-dependent support. |
 | `unseen_context` | boolean/string | Support for new cellular/patient contexts. |
 | `dose` | boolean/string | Whether dose is explicitly represented. |
 | `time` | boolean/string | Whether treatment time is explicitly represented. |
 | `patient_transfer` | boolean/string | Whether the method explicitly transfers to or predicts patient response. |
 | `code` | boolean/string | Whether an implementation is linked from the curated resource entry. |
 The Method Explorer is rendered at `docs/method-explorer.html`.
 ## Foundation Model Explorer profile
 Foundation models shown in the Foundation Model Explorer may define a `foundation_profile` object in a modular enrichment fragment.
 | Field | Type | Meaning |
 |---|---|---|
 | `year` | integer | Publication or release year used for navigation. |
 | `modalities` | array of strings | Fine-grained modalities such as `scrna`, `bulk-rna`, `proteomics`, `mutation`, `methylation`, `atac`, `spatial`, `pathology`, `dna`, `rna`, `chemical`, or `protein`. |
 | `params` | string or null | Human-readable parameter count or family range, e.g. `100M` or `1B / 7B / 40B`. |
 | `params_millions` | number | Largest representative parameter count in millions, used only for sorting. |
 | `pretraining_scale` | string or null | Human-readable summary of the pretraining corpus size. |
 | `species` | array of strings | Species represented in pretraining or primary use. |
 | `zero_shot` | boolean/string | Whether zero-shot use is explicitly supported; `limited` denotes task-dependent support. |
 | `finetunable` | boolean/string | Whether fine-tuning or parameter-efficient adaptation is supported. |
 | `weights` | boolean/string | Whether pretrained weights are publicly available. |
 | `code` | boolean/string | Whether implementation code is publicly available. |
 | `perturbation` | boolean/string | Whether perturbation prediction or in-silico perturbation is a supported use case. |
 | `spatial` | boolean/string | Whether spatial data are directly modeled or supported. |
 Unknown or unverified values should be omitted rather than inferred. The Foundation Model Explorer is rendered at `docs/foundation-explorer.html`.
 ## Agent Explorer profile
 Agentic AI systems shown in the Agent Explorer may define an `agent_profile` object in a modular enrichment fragment.
 | Field | Type | Meaning |
 |---|---|---|
 | `year` | integer | Initial public release or publication year. |
 | `domains` | array of strings | Main scientific domains such as `general-biomedicine`, `therapeutics`, `bioinformatics`, `single-cell`, `scientific-workflow`, or `paper-to-agent`. |
 | `architecture` | string | `single-agent`, `multi-agent`, or `agent-ecosystem`. |
 | `tool_use` | boolean/string | Whether external scientific tools/APIs are used. |
 | `code_execution` | boolean/string | Whether the system executes code as part of its workflow. |
 | `web_retrieval` | boolean/string | Whether web or online database retrieval is supported. |
 | `literature` | boolean/string | Whether literature search/reasoning is explicitly supported. |
 | `omics` | boolean/string | Whether omics analysis is a first-class supported workflow. |
 | `wet_lab` | boolean/string | Whether wet-lab planning, experimental design, or laboratory-facing workflows are supported. |
 | `autonomous_experiment` | boolean/string | Whether the system can carry out iterative computational/experimental research loops. |
 | `human_in_loop` | boolean/string | Whether human collaboration or approval is an explicit part of the design. |
 | `open_source` | boolean/string | Whether implementation code is publicly available. |
 Capability fields may use `limited` or `unknown` when support is workflow-dependent or not clearly documented. The Agent Explorer is rendered at `docs/agent-explorer.html`.