Cargando…

DNA sequence and structural properties as predictors of human and mouse promoters

Promoters play a central role in gene regulation, yet our power to discriminate them from non-promoter sequences in higher eukaryotes is mainly restricted to those associated with CpG islands. Here, we examined in silico the promoters of 30,954 human and 18,083 mouse transcripts in the DBTSS databas...

Descripción completa

Detalles Bibliográficos
Autores principales: Akan, Pelin, Deloukas, Panos
Formato: Texto
Lenguaje:English
Publicado: Elsevier/North-Holland 2008
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2672154/
https://www.ncbi.nlm.nih.gov/pubmed/18234453
http://dx.doi.org/10.1016/j.gene.2007.12.011
_version_ 1782166483960332288
author Akan, Pelin
Deloukas, Panos
author_facet Akan, Pelin
Deloukas, Panos
author_sort Akan, Pelin
collection PubMed
description Promoters play a central role in gene regulation, yet our power to discriminate them from non-promoter sequences in higher eukaryotes is mainly restricted to those associated with CpG islands. Here, we examined in silico the promoters of 30,954 human and 18,083 mouse transcripts in the DBTSS database, to assess the impact of particular sequence and structural features (propeller twist, bendability and nucleosome positioning preference) on promoter classification and prediction. Our analysis showed that a stricter-than-traditional definition of CpG islands captures low and high CpG count promoter classes more accurately than the traditional one. We observed that both human and mouse promoter sequences are flexible with the exception of the TATA box and TSS, which are rigid regions irrespective of association with a CpG island. Therefore varying levels of structural flexibility in promoters may affect their accessibility to proteins, and hence their specificity. For all features investigated, averaged values across core promoters discriminated CpG island associated promoters from background, whereas the same did not hold for promoters without a CpG island. However, local changes around − 34 to − 23 (expected position of TATA box) and the TSS were informative in discriminating promoters (both classes) from non-promoter sequences. Additionally, we investigated ATG deserts and observed that they occur in all promoter sets except those with a TATA-box and without a CpG island in human. Interestingly, all mouse promoter sets showed ATG codon depletion irrespective of the presence of a TATA-box, possibly reflecting a weaker contribution to TSS specificity in mouse.
format Text
id pubmed-2672154
institution National Center for Biotechnology Information
language English
publishDate 2008
publisher Elsevier/North-Holland
record_format MEDLINE/PubMed
spelling pubmed-26721542009-04-23 DNA sequence and structural properties as predictors of human and mouse promoters Akan, Pelin Deloukas, Panos Gene Article Promoters play a central role in gene regulation, yet our power to discriminate them from non-promoter sequences in higher eukaryotes is mainly restricted to those associated with CpG islands. Here, we examined in silico the promoters of 30,954 human and 18,083 mouse transcripts in the DBTSS database, to assess the impact of particular sequence and structural features (propeller twist, bendability and nucleosome positioning preference) on promoter classification and prediction. Our analysis showed that a stricter-than-traditional definition of CpG islands captures low and high CpG count promoter classes more accurately than the traditional one. We observed that both human and mouse promoter sequences are flexible with the exception of the TATA box and TSS, which are rigid regions irrespective of association with a CpG island. Therefore varying levels of structural flexibility in promoters may affect their accessibility to proteins, and hence their specificity. For all features investigated, averaged values across core promoters discriminated CpG island associated promoters from background, whereas the same did not hold for promoters without a CpG island. However, local changes around − 34 to − 23 (expected position of TATA box) and the TSS were informative in discriminating promoters (both classes) from non-promoter sequences. Additionally, we investigated ATG deserts and observed that they occur in all promoter sets except those with a TATA-box and without a CpG island in human. Interestingly, all mouse promoter sets showed ATG codon depletion irrespective of the presence of a TATA-box, possibly reflecting a weaker contribution to TSS specificity in mouse. Elsevier/North-Holland 2008-02-29 /pmc/articles/PMC2672154/ /pubmed/18234453 http://dx.doi.org/10.1016/j.gene.2007.12.011 Text en © 2008 Elsevier B.V. https://creativecommons.org/licenses/by/3.0/ Open Access under CC BY 3.0 (https://creativecommons.org/licenses/by/3.0/) license
spellingShingle Article
Akan, Pelin
Deloukas, Panos
DNA sequence and structural properties as predictors of human and mouse promoters
title DNA sequence and structural properties as predictors of human and mouse promoters
title_full DNA sequence and structural properties as predictors of human and mouse promoters
title_fullStr DNA sequence and structural properties as predictors of human and mouse promoters
title_full_unstemmed DNA sequence and structural properties as predictors of human and mouse promoters
title_short DNA sequence and structural properties as predictors of human and mouse promoters
title_sort dna sequence and structural properties as predictors of human and mouse promoters
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2672154/
https://www.ncbi.nlm.nih.gov/pubmed/18234453
http://dx.doi.org/10.1016/j.gene.2007.12.011
work_keys_str_mv AT akanpelin dnasequenceandstructuralpropertiesaspredictorsofhumanandmousepromoters
AT deloukaspanos dnasequenceandstructuralpropertiesaspredictorsofhumanandmousepromoters