Prediction of biological functions by histone modification patterns profiling

Yiou Xiao, Kishan G. Mehrotra, Chilukuri K. Mohan, Pratibha Choudhary, Ramesh Raina

Research output: Chapter in Book/Entry/PoemConference contribution

Abstract

Histone modifications provide an important layer of gene regulation in eukaryotes. In this paper, we propose an approach that identifies the histone modification patterns most relevant for specific biological functions, such as flowering in plants. We first propose a new pattern scoring method, which evaluates the importance of each combinatorial pattern of histone modifications; this is used along with logistic regression, Support Vector Machines, and naive Bayesian classifier algorithms to predict gene functions. This approach is shown to be successful in inferring significant patterns verified by independent gene function data, outperforming other pattern scores used in current histone modification analysis research.

Original languageEnglish (US)
Title of host publicationProceedings of the 9th International Conference on Bioinformatics and Computational Biology, BICOB 2017
EditorsHisham Al-Mubaid, Oliver Eulenstein, Qin Ding
PublisherThe International Society for Computers and Their Applications (ISCA)
Pages217-222
Number of pages6
ISBN (Electronic)9781943436071
StatePublished - 2017
Event9th International Conference on Bioinformatics and Computational Biology, BICOB 2017 - Honolulu, United States
Duration: Mar 20 2017Mar 22 2017

Publication series

NameProceedings of the 9th International Conference on Bioinformatics and Computational Biology, BICOB 2017

Other

Other9th International Conference on Bioinformatics and Computational Biology, BICOB 2017
Country/TerritoryUnited States
CityHonolulu
Period3/20/173/22/17

Keywords

  • Gene expression
  • Histone modification
  • Predicting biological function

ASJC Scopus subject areas

  • Biomedical Engineering
  • Health Information Management

Fingerprint

Dive into the research topics of 'Prediction of biological functions by histone modification patterns profiling'. Together they form a unique fingerprint.

Cite this