auto-optimize speeds up genomics tool leafcutter 10.1x with byte-identical output
jmschreiber91 · x · 2026-10-30
- A PR by jmschrei (UMass Chan's Programmable Genomics Laboratory) uses auto-optimize to speed up leafcutter's intron clustering and differential splicing: 100 GTEx v8 samples drop from 376.3s to 37.1s (10.1x faster, medians of five runs), with output byte-identical to the original.
- Across 93 Usage benchmark configurations (other cohorts, assays, assemblies, exon files, thread counts), median speedup is 2.6x and no run is more than 1s slower.
- The PR is split into three independently reviewable commits: clustering, differential splicing in R, and a compiled fast path for model fits.
- Notable as a case study of automated/AI-driven optimization of scientific computing code.
Related event: AI auto-optimize tool speeds up classic bioinformatics tools up to 10x(3 posts)→
More from Research
- Self-Direction turns agent mistakes into LoRA lessons, lifting GAIA success from 38.8% to 55.8% — rohanpaul_ai · 2026-10-30
- Preprint: AlphaGenome pervasively underestimates causal variant effects, undermining fine-mapping — anshulkundaje · 2026-10-30
- One Researcher on 22 NeurIPS 2026 Papers: What Should Author Lists Mean? — qberthet · 2026-10-30
- Stanford's I3L boosts whole-body robot deployment success by 68.3 points — jiajunwu_cs · 2026-10-30
- Dev ditches unanimous-vote stopping rule for a judge in multi-AI debate project — lluisantoni · 2026-10-30
- Google's AI Search Uses Your Knowledge Graph Facts Without Linking, RESONEO Research Finds — dejanseo · 2026-10-30