Modeling positional effects of regulatory sequences with spline transformations increases prediction accuracy of deep neural networks.

Journal: Bioinformatics (Oxford, England)
Published Date:

Abstract

MOTIVATION: Regulatory sequences are not solely defined by their nucleic acid sequence but also by their relative distances to genomic landmarks such as transcription start site, exon boundaries or polyadenylation site. Deep learning has become the approach of choice for modeling regulatory sequences because of its strength to learn complex sequence features. However, modeling relative distances to genomic landmarks in deep neural networks has not been addressed.

Authors

  • Žiga Avsec
    Department of Informatics, Technical University of Munich, 85748 Garching, Germany.
  • Mohammadamin Barekatain
    Department of Informatics, Technical University of Munich, 85748 Garching, Germany.
  • Jun Cheng
    School of Electrical and Information Technology, Yunnan Minzu University, Kunming, Yunnan 650500, PR China. Electronic address: jcheng6819@126.com.
  • Julien Gagneur
    Department of Informatics, Technical University of Munich, 85748 Garching, Germany.