Evaluating the informativeness of deep learning annotations for human complex diseases

Nat Commun. 2020 Sep 17;11(1):4703. doi: 10.1038/s41467-020-18515-4.


Deep learning models have shown great promise in predicting regulatory effects from DNA sequence, but their informativeness for human complex diseases is not fully understood. Here, we evaluate genome-wide SNP annotations from two previous deep learning models, DeepSEA and Basenji, by applying stratified LD score regression to 41 diseases and traits (average N = 320K), conditioning on a broad set of coding, conserved and regulatory annotations. We aggregated annotations across all (respectively blood or brain) tissues/cell-types in meta-analyses across all (respectively 11 blood or 8 brain) traits. The annotations were highly enriched for disease heritability, but produced only limited conditionally significant results: non-tissue-specific and brain-specific Basenji-H3K4me3 for all traits and brain traits respectively. We conclude that deep learning models have yet to achieve their full potential to provide considerable unique information for complex disease, and that their conditional informativeness for disease cannot be inferred from their accuracy in predicting regulatory annotations.

Publication types

  • Research Support, N.I.H., Extramural

MeSH terms

  • Alleles
  • Deep Learning*
  • Disease / genetics*
  • Genetic Predisposition to Disease
  • Genome, Human
  • Genome-Wide Association Study
  • Histones / genetics
  • Humans
  • Linkage Disequilibrium
  • Models, Genetic
  • Molecular Sequence Annotation*
  • Phenotype
  • Polymorphism, Single Nucleotide


  • Histones
  • histone H3 trimethyl Lys4