BEND
Human-genome tasks with biologically aligned splits; inspect each track for leakage controls.
7 benchmark-wide protocols · 105 benchmark-wide rows (not a task subtotal)D4 · Variant effect & evolution
Predict effects of non-coding variants on regulatory activity.
Benchmark coverage
Human-genome tasks with biologically aligned splits; inspect each track for leakage controls.
7 benchmark-wide protocols · 105 benchmark-wide rows (not a task subtotal)Regulatory DNA evaluation across zero-shot, probing and fine-tuning regimes.
0 benchmark-wide protocols · 0 benchmark-wide rows (not a task subtotal)One-time benchmark study rather than a maintained benchmark suite.
6 benchmark-wide protocols · 8 benchmark-wide rows (not a task subtotal)Only molecular and functional variant challenges are in scope; phenotype-from-exome and patient-level tasks are excluded.
0 benchmark-wide protocols · 0 benchmark-wide rows (not a task subtotal)Evaluation protocols
| Protocol | Track | Split / subset | Comparability |
|---|---|---|---|
P-B08-SQTL | Benchmarking DNA foundation models for genomic and genetic tasks — Regulatory variant-effect prediction B08-TR05 | nested chromosome-based held-out test groups putative causal sQTL | same_table_comparable cross-work compute/hardware equivalence |
P-B08-IPAQTL | Benchmarking DNA foundation models for genomic and genetic tasks — Regulatory variant-effect prediction B08-TR05 | nested chromosome-based held-out test groups putative causal ipaQTL | same_table_comparable cross-work compute/hardware equivalence |
P-B08-EQTL | Benchmarking DNA foundation models for genomic and genetic tasks — Regulatory variant-effect prediction B08-TR05 | nested chromosome-based held-out test groups putative causal eQTL | same_table_comparable cross-work compute/hardware equivalence |