Anthropic Computational Biology Eval
Private Anthropic computational-biology evaluation direction reported only through a model-trend chart, without public tasks, counts, or protocol details.
Scientific domain
Bioinformatics
Computational analysis of biological data and databases.
生物信息学
27 registered benchmarks and tracks
Anthropic Key Life Sciences Evals
An Anthropic private internal suite reported only through an official accuracy chart covering scientific figure interpretation, computational biology, and protein understanding.
BioMysteryBench
An agentic bioinformatics benchmark of objective, expert-authored mysteries over anonymized real-world biological data, scored on final answers rather than prescribed analysis paths.
BioSecBench-Surveillance
Agentic evaluation of pathogen genomic-surveillance workflow selection and analysis from raw or near-raw sequencing data.
BixBench
A containerized benchmark of long-horizon bioinformatics analysis over real published notebooks and associated data, with open-answer and multiple-choice evaluation modes.
CompBioBench
A 100-task agent benchmark of objectively gradable computational-biology problems requiring multi-step reasoning, bespoke code, tools, and real-world external resources.
crafted experiments
Real single-cell RNA-seq data augmented with known gene perturbations for comparing feature-selection methods.
GeneBench-Pro
A research-level agent benchmark of 129 synthetic, multistage computational-biology analyses that require iterative QC, statistical modeling, diagnostics, and decision-relevant judgment.
LAB-Bench
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.
LAB-Bench DbQA
Database-retrieval category spanning 10 genomics, clinical, protein, regulatory, vaccine-response, and viral-PPI tasks.
LAB-Bench DbQA — Disease gene associations
Identifies genes associated with a phenotype in DisGeNET but not OMIM.
LAB-Bench DbQA — Gene location
Retrieves human-gene cytogenetic locations from the stated Ensembl release.
LAB-Bench DbQA — miRNA targets
Retrieves computationally predicted human miRNA targets from miRDB.
LAB-Bench DbQA — Mouse tumor gene sets
Retrieves genes in Mammalian Phenotype Tumor Ontology gene sets.
LAB-Bench DbQA — Oncogenic signatures
Retrieves membership in MSigDB C6 oncogenic-signature gene sets.
LAB-Bench DbQA — GTRD transcription-factor binding sites
Retrieves promoter-region transcription-factor binding-site annotations from GTRD.
LAB-Bench DbQA — Protein variant from sequence
Uses a protein sequence and ClinVar lookup to identify benign or pathogenic variants.
LAB-Bench DbQA — Protein variant with multiple sequences
Identifies ClinVar variant pathogenicity while reasoning across multiple protein sequences.
LAB-Bench DbQA — Vaccine response gene sets
Retrieves membership in MSigDB vaccine-response gene sets.
LAB-Bench DbQA — Viral protein–protein interactions
Retrieves predicted human interaction partners of viral proteins from P-HIPSter.
LifeSciBench
Expert-authored, artifact-rich free-response tasks that evaluate realistic research judgment across applied life-science workflows.
scBench
Agentic evaluation suite for data-grounded single-cell analysis across diverse sequencing technologies and workflow stages.
SCIGYM
An agentic systems-biology suite in which language models iteratively perturb simulated SBML systems, analyze time-series observations in Python, and reconstruct hidden biological reactions.
SCIGYM Large
The formally released SCIGYM track containing the 213 systems not included in the creator paper's model evaluation, with systems reaching up to 400 reactions.
SCIGYM Small
The formally released and creator-evaluated SCIGYM track containing biological systems with fewer than ten reactions.
SpatialBench
A benchmark of deterministic, verifiable agentic problems derived from real spatial-transcriptomics workflows, testing whether agents can manipulate data and recover key biological results.
VirBench
Retrieval benchmark that tests whether scientific agents can answer verified viral-sequence questions by querying NCBI Virus.
Capability coverage
Registry records tagged Bioinformatics, counted by capability.
Accessible data table
| Capability | Records |
|---|---|
| Knowledge | 11 |
| Evidence synthesis | 3 |
| Retrieval | 15 |
| Prediction | 4 |
| Classification | 4 |
| Design | 2 |
| Generation | 1 |
| Optimization | 1 |
| Data analysis | 16 |
| Coding | 9 |
| Tool use | 12 |
| Experiment planning | 5 |
| Troubleshooting | 2 |
| Scientific reasoning | 18 |
| Scientific communication | 1 |
Scientific Tasks in this domain
Task mappings are evidence-backed and may be partial for mixed suites.
Coverage-gap reading: zero counts indicate a gap in this registry, not proof that no benchmark exists. Propose a primary source through the contribution form.