2007•GeneOpen access

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

Pelin Akan, Panos Deloukas

Open full text 48 citations

Abstract

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.

Open-access reader

About this research paper

What this paper is about

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.

Why it matters

OpenAlex reports 48 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

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.

Key concepts: Biology, Promoter, DNA, Sequence (biology), Genetics, DNA sequencing, Recombinant DNA, Computational biology

Related papers

Back to paper searchBrowse research topicsOriginal source
DNA sequence and structural properties as predictors of human and mouse promoters — Research Paper | ScholarLens