Rxivist logo

Rxivist combines preprints from bioRxiv with data from Twitter to help you find the papers being discussed in your field. Currently indexing 73,530 bioRxiv papers from 320,027 authors.

Reverse-complement parameter sharing improves deep learning models for genomics

By Avanti Shrikumar, Peyton Greenside, Anshul Kundaje

Posted 27 Jan 2017
bioRxiv DOI: 10.1101/103663

Deep learning approaches that have produced breakthrough predictive models in computer vision, speech recognition and machine translation are now being successfully applied to problems in regulatory genomics. However, deep learning architectures used thus far in genomics are often directly ported from computer vision and natural language processing applications with few, if any, domain-specific modifications. In double-stranded DNA, the same pattern may appear identically on one strand and its reverse complement due to complementary base pairing. Here, we show that conventional deep learning models that do not explicitly model this property can produce substantially different predictions on forward and reverse-complement versions of the same DNA sequence. We present four new convolutional neural network layers that leverage the reverse-complement property of genomic DNA sequence by sharing parameters between forward and reverse-complement representations in the model. These layers guarantee that forward and reverse-complement sequences produce identical predictions within numerical precision. Using experiments on simulated and in vivo transcription factor binding data, we show that our proposed architectures lead to improved performance, faster learning and cleaner internal representations compared to conventional architectures trained on the same data.

Download data

  • Downloaded 3,097 times
  • Download rankings, all-time:
    • Site-wide: 1,256 out of 73,530
    • In bioinformatics: 249 out of 7,162
  • Year to date:
    • Site-wide: 5,675 out of 73,530
  • Since beginning of last month:
    • Site-wide: 5,675 out of 73,530

Altmetric data

Downloads over time

Distribution of downloads per paper, site-wide


Sign up for the Rxivist weekly newsletter! (Click here for more details.)