The more papers I read for a review article I'm writing about ML pitfalls in genomics, the more my faith is shaken in the results from papers that apply machine learning to methylation arrays. A salty thread. 1/
As the last drop of the week, we're releasing the Cherimoya Binding Atlas (BATv1), a collection of genomic S2F models for 3,429 TF binding ChIP-seq experiments from ENCODE covering 183 biosamples and 1,139 distinct targets.
Time for the biggest drop of the week: auto-optimizations of eleven bioinformatics packages, focusing on read alignment. Each auto-optimization includes: a new unit test suite, the optimizations stacked on top, and a list of issues found during the validation.
I've always had trouble mapping genomic motifs <-> TFs because many TFs can bind a motif. But.. this is important for understanding genomic S2F models and designed DNA. As my third drop, I'm releasing a Claude artifact for exploring these connections.