Discovering sequences with potential regulatory characteristics.

Journal Article (Journal Article)

We developed a computational model to explore the hypothesis that regulatory instructions are context dependent and conveyed through specific 'codes' in human genomic DNA. We provide examples of correlation of computational predictions to reported mapped DNase I hypersensitive segments in the HOXA locus in human chromosome 7. The examples show that statistically significant 9-mers from promoter regions may occur in sequences near and upstream of transcription initiation sites, in intronic regions, and within intergenic regions. Additionally, a subset of 9-mers from coding sequences appears frequently, as clusters, in regulatory regions dispersed in noncoding regions in genomic DNA. The results suggest that the computational model has the potential of decoding regulatory instructions to discover candidate transcription factor binding sites and to discover candidate epigenetic signals that appear in both coding and regulatory regions of genes.

Full Text

Duke Authors

Cited Authors

  • Bina, M; Wyss, P; Lazarus, SA; Shah, SR; Ren, W; Szpankowski, W; Crawford, GE; Park, SP; Song, XC

Published Date

  • April 2009

Published In

Volume / Issue

  • 93 / 4

Start / End Page

  • 314 - 322

PubMed ID

  • 19084590

Electronic International Standard Serial Number (EISSN)

  • 1089-8646

Digital Object Identifier (DOI)

  • 10.1016/j.ygeno.2008.11.008


  • eng

Conference Location

  • United States