Calico open-sources Cerberus DNA model in PyTorch: 786kb input, 8361 human tracks
anshulkundaje · x · 2026-10-10
Calico released the baskerville repo with Apache 2.0 PyTorch models of Cerberus and predecessors Borzoi/Borzoi Prime, predicting regulatory activity (chromatin accessibility, gene expression) from DNA sequence.
- Cerberus swaps Borzoi's transformer for bidirectional state-space (Hydra) blocks, reads 786 kb of sequence, and predicts 8,361 human and 3,102 mouse functional genomics tracks at 32 bp resolution (RNA-seq, CAGE, DNase/ATAC, ChIP, CLIP)
- The group moved its software infrastructure from TensorFlow to PyTorch
- Cerberus is reported more accurate and cheaper to run than Borzoi
More from Research
- Elliot Glazer on when a Lean proof actually proves the math we care about — ctjlewis · 2026-10-10
- CS theorist Lance Fortnow on AI and math: don't panic, read a new proof — fortnow · 2026-10-10
- Skill Constellations: first dated copy network of 2.1M agent-skill adoptions on GitHub — UBC-O · 2026-10-10
- Physical attention bias cuts cable-simulation prediction error by 15%+ — Avihai Giuili · 2026-10-10
- GitSwarm paper: agents collaborate through a shared Git repo, solving all 30 IMOProofBench-Advanced problems — anirudhg9119 · 2026-10-10
- GitSwarm agents spontaneously adapt their workflow: building on coding tasks, verifying on proof tasks — anirudhg9119 · 2026-10-10