Cargando…
UDSMProt: universal deep sequence models for protein classification
MOTIVATION: Inferring the properties of a protein from its amino acid sequence is one of the key problems in bioinformatics. Most state-of-the-art approaches for protein classification are tailored to single classification tasks and rely on handcrafted features, such as position-specific-scoring mat...
Autores principales: | , , , |
---|---|
Formato: | Online Artículo Texto |
Lenguaje: | English |
Publicado: |
Oxford University Press
2020
|
Materias: | |
Acceso en línea: | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7178389/ https://www.ncbi.nlm.nih.gov/pubmed/31913448 http://dx.doi.org/10.1093/bioinformatics/btaa003 |
_version_ | 1783525445808947200 |
---|---|
author | Strodthoff, Nils Wagner, Patrick Wenzel, Markus Samek, Wojciech |
author_facet | Strodthoff, Nils Wagner, Patrick Wenzel, Markus Samek, Wojciech |
author_sort | Strodthoff, Nils |
collection | PubMed |
description | MOTIVATION: Inferring the properties of a protein from its amino acid sequence is one of the key problems in bioinformatics. Most state-of-the-art approaches for protein classification are tailored to single classification tasks and rely on handcrafted features, such as position-specific-scoring matrices from expensive database searches. We argue that this level of performance can be reached or even be surpassed by learning a task-agnostic representation once, using self-supervised language modeling, and transferring it to specific tasks by a simple fine-tuning step. RESULTS: We put forward a universal deep sequence model that is pre-trained on unlabeled protein sequences from Swiss-Prot and fine-tuned on protein classification tasks. We apply it to three prototypical tasks, namely enzyme class prediction, gene ontology prediction and remote homology and fold detection. The proposed method performs on par with state-of-the-art algorithms that were tailored to these specific tasks or, for two out of three tasks, even outperforms them. These results stress the possibility of inferring protein properties from the sequence alone and, on more general grounds, the prospects of modern natural language processing methods in omics. Moreover, we illustrate the prospects for explainable machine learning methods in this field by selected case studies. AVAILABILITY AND IMPLEMENTATION: Source code is available under https://github.com/nstrodt/UDSMProt. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online. |
format | Online Article Text |
id | pubmed-7178389 |
institution | National Center for Biotechnology Information |
language | English |
publishDate | 2020 |
publisher | Oxford University Press |
record_format | MEDLINE/PubMed |
spelling | pubmed-71783892020-04-28 UDSMProt: universal deep sequence models for protein classification Strodthoff, Nils Wagner, Patrick Wenzel, Markus Samek, Wojciech Bioinformatics Original Papers MOTIVATION: Inferring the properties of a protein from its amino acid sequence is one of the key problems in bioinformatics. Most state-of-the-art approaches for protein classification are tailored to single classification tasks and rely on handcrafted features, such as position-specific-scoring matrices from expensive database searches. We argue that this level of performance can be reached or even be surpassed by learning a task-agnostic representation once, using self-supervised language modeling, and transferring it to specific tasks by a simple fine-tuning step. RESULTS: We put forward a universal deep sequence model that is pre-trained on unlabeled protein sequences from Swiss-Prot and fine-tuned on protein classification tasks. We apply it to three prototypical tasks, namely enzyme class prediction, gene ontology prediction and remote homology and fold detection. The proposed method performs on par with state-of-the-art algorithms that were tailored to these specific tasks or, for two out of three tasks, even outperforms them. These results stress the possibility of inferring protein properties from the sequence alone and, on more general grounds, the prospects of modern natural language processing methods in omics. Moreover, we illustrate the prospects for explainable machine learning methods in this field by selected case studies. AVAILABILITY AND IMPLEMENTATION: Source code is available under https://github.com/nstrodt/UDSMProt. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online. Oxford University Press 2020-04-15 2020-01-08 /pmc/articles/PMC7178389/ /pubmed/31913448 http://dx.doi.org/10.1093/bioinformatics/btaa003 Text en © The Author(s) 2020. Published by Oxford University Press. http://creativecommons.org/licenses/by/4.0/ This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. |
spellingShingle | Original Papers Strodthoff, Nils Wagner, Patrick Wenzel, Markus Samek, Wojciech UDSMProt: universal deep sequence models for protein classification |
title | UDSMProt: universal deep sequence models for protein classification |
title_full | UDSMProt: universal deep sequence models for protein classification |
title_fullStr | UDSMProt: universal deep sequence models for protein classification |
title_full_unstemmed | UDSMProt: universal deep sequence models for protein classification |
title_short | UDSMProt: universal deep sequence models for protein classification |
title_sort | udsmprot: universal deep sequence models for protein classification |
topic | Original Papers |
url | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7178389/ https://www.ncbi.nlm.nih.gov/pubmed/31913448 http://dx.doi.org/10.1093/bioinformatics/btaa003 |
work_keys_str_mv | AT strodthoffnils udsmprotuniversaldeepsequencemodelsforproteinclassification AT wagnerpatrick udsmprotuniversaldeepsequencemodelsforproteinclassification AT wenzelmarkus udsmprotuniversaldeepsequencemodelsforproteinclassification AT samekwojciech udsmprotuniversaldeepsequencemodelsforproteinclassification |