domains / families / science / deep-learning-atac

deep-learning-atac

separate

Sequence-based deep learning for ATAC-seq using chromBPNet, BPNet, scBasset, or Enformer. Use when correcting Tn5 bias with neural networks beyond k-mer models, predicting per-base accessibility profiles, scoring in silico variant effects at GWAS or rare-variant SNPs, discovering motifs via DeepLIFT/TF-MoDISco from a trained model, or generating cell-type-specific accessibility predictions for unobserved cell states.

When to use

• correcting Tn5 bias with neural networks
• predicting per-base accessibility profiles
• scoring in silico variant effects at GWAS or rare-variant SNPs
• discovering motifs via DeepLIFT/TF-MoDISco
• generating cell-type-specific accessibility predictions
• per-cell sequence-based predictions
• long-context epigenome prediction
• variant effect prediction in non-genic regions

When not to use

• standard peak calling
• differential accessibility analysis
• replacement for MACS+TOBIAS at every step
• new work using DeepATAC or Basset

Skill page

Source