Selene: a PyTorch-based deep learning library for sequence data

Kathleen M Chen; Evan M Cofer; Jian Zhou; Olga G Troyanskaya

doi:10.1038/s41592-019-0360-8

Selene: a PyTorch-based deep learning library for sequence data

Nat Methods. 2019 Apr;16(4):315-318. doi: 10.1038/s41592-019-0360-8. Epub 2019 Mar 28.

Authors

Kathleen M Chen¹, Evan M Cofer^{2

3}, Jian Zhou^{1

2}, Olga G Troyanskaya^{4

5

6}

Affiliations

¹ Flatiron Institute, Simons Foundation, New York, NY, USA.
² Lewis-Sigler Institute for Integrative Genomics, Princeton University, Princeton, NJ, USA.
³ Graduate Program in Quantitative and Computational Biology, Princeton University, Princeton, NJ, USA.
⁴ Flatiron Institute, Simons Foundation, New York, NY, USA. ogt@cs.princeton.edu.
⁵ Lewis-Sigler Institute for Integrative Genomics, Princeton University, Princeton, NJ, USA. ogt@cs.princeton.edu.
⁶ Department of Computer Science, Princeton University, Princeton, NJ, USA. ogt@cs.princeton.edu.

Abstract

To enable the application of deep learning in biology, we present Selene (https://selene.flatironinstitute.org/), a PyTorch-based deep learning library for fast and easy development, training, and application of deep learning model architectures for any biological sequence data. We demonstrate on DNA sequences how Selene allows researchers to easily train a published architecture on new data, develop and evaluate a new architecture, and use a trained model to answer biological questions of interest.

Publication types

Research Support, N.I.H., Extramural
Research Support, Non-U.S. Gov't

MeSH terms

Algorithms
Alzheimer Disease / metabolism
Area Under Curve
Computational Biology / methods*
Deep Learning*
Gene Library
Genomics
Humans
Models, Statistical
Mutagenesis
Mutation
Neural Networks, Computer*
Normal Distribution
Programming Languages
Sequence Analysis, DNA*
Software

Abstract

Publication types

MeSH terms

Grants and funding