Cargando…

Extracting sequence features to predict protein–DNA interactions: a comparative study

Predicting how and where proteins, especially transcription factors (TFs), interact with DNA is an important problem in biology. We present here a systematic study of predictive modeling approaches to the TF–DNA binding problem, which have been frequently shown to be more efficient than those method...

Descripción completa

Detalles Bibliográficos
Autores principales: Zhou, Qing, Liu, Jun S.
Formato: Texto
Lenguaje:English
Publicado: Oxford University Press 2008
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2475627/
https://www.ncbi.nlm.nih.gov/pubmed/18556756
http://dx.doi.org/10.1093/nar/gkn361
_version_ 1782157565225861120
author Zhou, Qing
Liu, Jun S.
author_facet Zhou, Qing
Liu, Jun S.
author_sort Zhou, Qing
collection PubMed
description Predicting how and where proteins, especially transcription factors (TFs), interact with DNA is an important problem in biology. We present here a systematic study of predictive modeling approaches to the TF–DNA binding problem, which have been frequently shown to be more efficient than those methods only based on position-specific weight matrices (PWMs). In these approaches, a statistical relationship between genomic sequences and gene expression or ChIP-binding intensities is inferred through a regression framework; and influential sequence features are identified by variable selection. We examine a few state-of-the-art learning methods including stepwise linear regression, multivariate adaptive regression splines, neural networks, support vector machines, boosting and Bayesian additive regression trees (BART). These methods are applied to both simulated datasets and two whole-genome ChIP-chip datasets on the TFs Oct4 and Sox2, respectively, in human embryonic stem cells. We find that, with proper learning methods, predictive modeling approaches can significantly improve the predictive power and identify more biologically interesting features, such as TF–TF interactions, than the PWM approach. In particular, BART and boosting show the best and the most robust overall performance among all the methods.
format Text
id pubmed-2475627
institution National Center for Biotechnology Information
language English
publishDate 2008
publisher Oxford University Press
record_format MEDLINE/PubMed
spelling pubmed-24756272008-07-21 Extracting sequence features to predict protein–DNA interactions: a comparative study Zhou, Qing Liu, Jun S. Nucleic Acids Res Computational Biology Predicting how and where proteins, especially transcription factors (TFs), interact with DNA is an important problem in biology. We present here a systematic study of predictive modeling approaches to the TF–DNA binding problem, which have been frequently shown to be more efficient than those methods only based on position-specific weight matrices (PWMs). In these approaches, a statistical relationship between genomic sequences and gene expression or ChIP-binding intensities is inferred through a regression framework; and influential sequence features are identified by variable selection. We examine a few state-of-the-art learning methods including stepwise linear regression, multivariate adaptive regression splines, neural networks, support vector machines, boosting and Bayesian additive regression trees (BART). These methods are applied to both simulated datasets and two whole-genome ChIP-chip datasets on the TFs Oct4 and Sox2, respectively, in human embryonic stem cells. We find that, with proper learning methods, predictive modeling approaches can significantly improve the predictive power and identify more biologically interesting features, such as TF–TF interactions, than the PWM approach. In particular, BART and boosting show the best and the most robust overall performance among all the methods. Oxford University Press 2008-07 2008-06-13 /pmc/articles/PMC2475627/ /pubmed/18556756 http://dx.doi.org/10.1093/nar/gkn361 Text en © 2008 The Author(s) http://creativecommons.org/licenses/by-nc/2.0/uk/ This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
spellingShingle Computational Biology
Zhou, Qing
Liu, Jun S.
Extracting sequence features to predict protein–DNA interactions: a comparative study
title Extracting sequence features to predict protein–DNA interactions: a comparative study
title_full Extracting sequence features to predict protein–DNA interactions: a comparative study
title_fullStr Extracting sequence features to predict protein–DNA interactions: a comparative study
title_full_unstemmed Extracting sequence features to predict protein–DNA interactions: a comparative study
title_short Extracting sequence features to predict protein–DNA interactions: a comparative study
title_sort extracting sequence features to predict protein–dna interactions: a comparative study
topic Computational Biology
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2475627/
https://www.ncbi.nlm.nih.gov/pubmed/18556756
http://dx.doi.org/10.1093/nar/gkn361
work_keys_str_mv AT zhouqing extractingsequencefeaturestopredictproteindnainteractionsacomparativestudy
AT liujuns extractingsequencefeaturestopredictproteindnainteractionsacomparativestudy