cyberneticlibrary

Train genomic region embeddings

genimlskillsetup L2★27,559

K-Dense-AI/scientific-agent-skills ↗

What it does

Train machine learning models on genomic regions

Best for

Unsupervised learning on genomic interval data, region embeddings, and single-cell ATAC analysis.

Inputs

· BED file collection
· Universe peak reference
· Optional: cell metadata

Outputs

· Region embeddings array
· Model weights
· Cluster assignments

Requires

· PyTorch
· Word2vec/StarSpace
· scanpy optional

Preconditions

· BED format valid
· Universe prebuilt

Failure modes

· Insufficient coverage
· Tokenization mismatch
· OOM

Trust signals

· BSD-2-Clause license
· databio/geniml repo