Biology-Instructions
A multi-omics sequence-instruction suite with 21 formal predictive tasks across DNA, RNA, protein, and multi-molecule inputs; the final paper reports task-specific held-out evaluations rather than a single aggregate score.
Scientific task · Molecular interaction
Predict or analyze interactions
分子相互作用与结合
molecular-interaction-analysisCoverage
A multi-omics sequence-instruction suite with 21 formal predictive tasks across DNA, RNA, protein, and multi-molecule inputs; the final paper reports task-specific held-out evaluations rather than a single aggregate score.
Binary neutralization prediction for an antibody-antigen protein-sequence pair, evaluated with Matthews correlation coefficient.
Dedicated CASP17 category for blind prediction of antibody-antigen, nanobody-antigen, and T-cell receptor complex structures.
A creator-curated set of 944 protein-science multiple-choice questions with answer explanations, generated from research literature and released for evaluating text LLM protein understanding.
A living collection of eight curated 3D molecular-learning tasks spanning small molecules, protein interactions and mutations, ligand binding, and protein/RNA structure ranking.
Biennial blind community experiments that assess macromolecular structure, complex, ligand, and model-accuracy prediction against experimental structures withheld during prediction.
Formal CASP track for blind prediction of protein-ligand binding poses, binding affinity or rank, binding pockets, and pose confidence.
The original molecular-machine-learning benchmark of 17 dataset collections and more than 800 prediction endpoints spanning quantum, physicochemical, biophysical, and physiological properties.
Versioned deep-mutational-scanning and clinical-variant benchmarks for protein fitness prediction and design in zero-shot and supervised regimes.
Weekly, automated, independent, blind evaluation of registered macromolecular structure-prediction servers on complete PDB entries whose experimental structures are withheld during prediction.
A practical biology-research suite of 2,457 multiple-choice questions across eight broad categories and 31 versioned task files, with public and private contamination-monitoring splits.
Retrieves predicted human interaction partners of viral proteins from P-HIPSter.
Binary interaction prediction for an RNA and protein sequence pair, evaluated with Matthews correlation coefficient.
Each row keeps its original unit and basis. Rows with different units or overlapping mappings are never added.
| Benchmark | Mapped task | Coverage | Count | Version | Evidence |
|---|---|---|---|---|---|
| Biology-Instructions root: bioinstruction | Antibody-antigen interaction official-track · high | explicitly-in-scope | 1 tracks Formal evaluation tracks (Antibody-Antigen Neutralization). | emnlp-2025 | bioinstruction-evidence-paper |
| Biology-Instructions Antibody-Antigen Neutralization root: bioinstruction | Antibody-antigen interaction official-track · high | explicitly-in-scope | 26902 examples distinct examples across the published train, validation, and test splits | emnlp-2025 | bioinstruction-aan-evidence-paper |
| CASP17 Immune Complexes root: casp | Antibody-antigen interaction official-track · high | explicitly-in-scope | Not reported CASP17 immune-complex targets. | CASP17 as of 2026-07-21 | casp-immune-evidence-category |
| ProteinLMBench root: proteinlmbench | Molecular interaction and binding official-taxonomy · high | observed | Not reported Released ProteinLMBench question records. | hf-f139796 | proteinlmbench-evidence-paper |
| ATOM3D root: atom3d | Protein-complex mutation stability prediction official-track · high | explicitly-in-scope | 1 tasks Curated 3D benchmark datasets. | v0.2.6 as of 2026-07-22 | atom3d-paper-definition-evidence |
| ATOM3D root: atom3d | Protein-ligand binding affinity official-track · high | explicitly-in-scope | 1 tasks Curated 3D benchmark datasets. | v0.2.6 as of 2026-07-22 | atom3d-paper-definition-evidence |
| CASP root: casp | Protein-ligand binding affinity official-track · high | explicitly-in-scope | Not reported CASP17 protein-ligand targets. | CASP17 as of 2026-07-21 | casp-evidence-casp17-protocol |
| CASP Protein-Ligand Prediction root: casp | Protein-ligand binding affinity official-track · high | explicitly-in-scope | Not reported CASP17 protein-ligand targets. | CASP17 as of 2026-07-21 | casp-ligand-evidence-casp17 |
| MoleculeNet root: moleculenet | Protein-ligand binding affinity official-taxonomy · high | explicitly-in-scope | 1 other Original paper dataset collections. | original-2017 | moleculenet-paper-definition-evidence |
| LifeSciBench root: lifescibench | Protein-ligand binding prediction official-taxonomy · high | explicitly-in-scope | Not reported Expert-authored benchmark tasks. | initial-release | lifescibench-evidence-taxonomy |
| ProteinGym root: proteingym | Protein-ligand binding prediction official-taxonomy · high | observed | Not reported DMS assays in the official generic Binding function category. | 1.3 | proteingym-evidence-taxonomy |
| ATOM3D root: atom3d | Protein-ligand functional efficacy prediction official-track · high | explicitly-in-scope | 1 tasks Curated 3D benchmark datasets. | v0.2.6 as of 2026-07-22 | atom3d-paper-definition-evidence |
| CAMEO root: cameo | Protein-ligand pose prediction official-taxonomy · high | explicitly-in-scope | Not reported Ligand-containing targets in the current rolling service. | current-complex-3d as of 2026-07-21 | cameo-evidence-2024-study |
| CASP root: casp | Protein-ligand pose prediction official-track · high | explicitly-in-scope | Not reported CASP17 protein-ligand targets. | CASP17 as of 2026-07-21 | casp-evidence-casp17-protocol |
| CASP Protein-Ligand Prediction root: casp | Protein-ligand pose prediction official-track · high | explicitly-in-scope | Not reported CASP17 protein-ligand targets. | CASP17 as of 2026-07-21 | casp-ligand-evidence-casp17 |
| LAB-Bench root: lab-bench | Protein-protein interaction prediction official-track · high | explicitly-in-scope | 50 questions Viral PPI formal-task questions. | repository-998a8e0 | lab-bench-evidence-paper |
| LAB-Bench DbQA — Viral protein–protein interactions root: lab-bench | Protein-protein interaction prediction official-track · high | explicitly-in-scope | 50 questions questions across public and private splits | repository-998a8e0 | lab-bench-dbqa-viral-ppi-evidence-paper |
| LifeSciBench root: lifescibench | Protein-protein interaction prediction official-taxonomy · high | explicitly-in-scope | Not reported Expert-authored benchmark tasks. | initial-release | lifescibench-evidence-taxonomy |
| ATOM3D root: atom3d | Protein-protein interface prediction official-track · high | explicitly-in-scope | 1 tasks Curated 3D benchmark datasets. | v0.2.6 as of 2026-07-22 | atom3d-paper-definition-evidence |
| Biology-Instructions root: bioinstruction | RNA-protein interaction prediction official-track · high | explicitly-in-scope | 1 tracks Formal evaluation tracks (RNA-Protein Interaction Prediction). | emnlp-2025 | bioinstruction-evidence-paper |
| Biology-Instructions RNA-Protein Interaction Prediction root: bioinstruction | RNA-protein interaction prediction official-track · high | explicitly-in-scope | 20824 examples distinct examples across the published train, validation, and test splits | emnlp-2025 | bioinstruction-rpi-evidence-paper |
Runs are included only for benchmark records mapped here (and formal child tracks when a mapped suite is the root). A task mapping does not imply that every run isolates this task.