Sequential regulatory activity prediction across chromosomes with convolutional neural networks

Models for predicting phenotypic outcomes from genotypes have important applications to understanding genomic function and improving human health. Here, we develop a machine-learning system to predict cell-type-specific epigenetic and transcriptional profiles in large mammalian genomes from DNA sequ...

Full description

Saved in:
Bibliographic Details
Published inGenome research Vol. 28; no. 5; pp. 739 - 750
Main Authors Kelley, David R, Reshef, Yakir A, Bileschi, Maxwell, Belanger, David, McLean, Cory Y, Snoek, Jasper
Format Journal Article
LanguageEnglish
Published United States Cold Spring Harbor Laboratory Press 01.05.2018
Subjects
Chromosomes
Epigenetics
Gene expression
Gene mapping
Genotypes
Learning algorithms
Method
Neural networks
Nucleotide sequence
Predictions
Regulatory sequences
Transcription
Online AccessGet full text

Cover

More Information
Summary:Models for predicting phenotypic outcomes from genotypes have important applications to understanding genomic function and improving human health. Here, we develop a machine-learning system to predict cell-type-specific epigenetic and transcriptional profiles in large mammalian genomes from DNA sequence alone. By use of convolutional neural networks, this system identifies promoters and distal regulatory elements and synthesizes their content to make effective gene expression predictions. We show that model predictions for the influence of genomic variants on gene expression align well to causal variants underlying eQTLs in human populations and can be useful for generating mechanistic hypotheses to enable fine mapping of disease loci.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ISSN:1088-9051
1549-5469
DOI:10.1101/gr.227819.117