A large-scale analysis for significance assessment of frequencies relative to potentially strong sigma 70 promoters - comparison of 32 prokaryotic genomes - - LINA - Equipe COMbinatoire et Bio-Informatique Accéder directement au contenu
Pré-Publication, Document De Travail Année : 2007

A large-scale analysis for significance assessment of frequencies relative to potentially strong sigma 70 promoters - comparison of 32 prokaryotic genomes -

Résumé

This report presents a computational analysis of high ORF expression potentialities in prokaryotic genomes, in relation with medical and economic relevance. Given a bacterial genome and the description of a structured motif, the software BacTrans2 implements the search of the occurrence most similar to the motif, in the regulatory region of each gene. In this work, the software BacTrans2 was run over 32 prokaryotic genomes, to identify putative strong sigma 70 promoters. We focused in particular on sigma 70 promoters harbouring an UP element, which enhances transcription initiation. We performed four computational analyses per genome, combining two promoter strength levels (CI > CII) with either mandatory or optional UP element presence. We compared the frequencies obtained for 32 bacterial genomes, under these four constraint specifications. First, we show that an over-representation of putative strong promoters differentiates the AT-rich Firmicutes' genomes from other genomes. Another interesting result is that strong promoters of relatively lesser quality (CII) are more frequently associated with an UP element than strong promoters of better quality (CI). Then, per each bacterial genome studied, we generated at random 100 artificial genomes. Such genomes are only constrained as to have the same two following characteristics as the bacterial genome: same total number of genes and same proportions of A, C, T and G nucleotides in the 350 nucleotide-long region upstream of start codon. The sigma 70 promoter frequency observed on average over these 100 genomes is compared to the frequency observed for the bacterial genome, under each of the four constraint sets aforementioned. Thus, the statistical significance of the sigma 70 model is discussed for each genome, under each constraint set. For most genomes, and especially for Firmicutes, a meaningful difference is statistically ascertained. Besides, the comparison between Firmicutes genomes and equally AT-rich Proteobacteria genomes also confirm that the Firmicutes specificity is not related to genome size bias. Hence, Firmicutes would appear as genomes more favoured by nature with respect to high intrinsic transcription potentiality. Throughout the report, we discuss the influence of AT-richness on promoter frequencies, implementing various correlation analyses. We show that an influence is only observed when the UP element is required. Then we evaluate whether the statistical significance of the sigma 70 model is related or not to AT-richness. Interestingly, we find that the relation is loose except when the UP element is required, and under the more stringent constraint (CI). Thus, we distinguish the AT-bias, whose influence is more or less noticeable for bacterial genomes as well as randomly generated genomes, whatever the species, and the species bias such as the one identified for Firmicutes. Finally, we compare the AT-percentages of three sub-regions of the 350 bp-long region upstream of start codon, distinguishing between genes harbouring strong promoters and genes not harbouring any such strong promoters. We can show no evidence that the over-representation characterizing Firmicutes is due to a local AT-bias. To our knowledge, our work is the very first statistical approach thoroughly analysing the presence significance of various potentially strong sigma 70 promoter models, including models harbouring the UP element enhancer, in the context of a genome-comparative study. The presence of the enhanced promoter has been proven significant in all eight large Firmicutes genomes studied and between ten and thirteen non Firmicutes large genomes studied (depending on the Z-score threshold considered).
Fichier principal
Vignette du fichier
research_report_nar_tues_12_06_07.pdf (496.62 Ko) Télécharger le fichier
appendix1_sinoquet_wed_13_06_07.pdf (41.32 Ko) Télécharger le fichier
appendix2_1_sinoquet_et_al.html (90.02 Ko) Télécharger le fichier
appendix2_2_sinoquet_et_al.html (235.8 Ko) Télécharger le fichier
appendix3_1_sinoquet_et_al.html (4.07 Ko) Télécharger le fichier
appendix3_2_sinoquet_et_al.html (26.49 Ko) Télécharger le fichier
appendix4_sinoquet_wed_13_06_07.pdf (33.16 Ko) Télécharger le fichier
appendix5_sinoquet_wed_13_06_07.pdf (41.71 Ko) Télécharger le fichier
appendix6_sinoquet_wed_13_06_07.pdf (38.16 Ko) Télécharger le fichier
appendix7_sinoquet_wed_13_06_07.pdf (30.65 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Format : Autre
Format : Autre
Format : Autre
Format : Autre
Format : Autre
Format : Autre
Format : Autre
Format : Autre
Format : Autre

Dates et versions

hal-00153303 , version 1 (14-06-2007)
hal-00153303 , version 2 (26-07-2007)
hal-00153303 , version 3 (30-10-2007)

Identifiants

  • HAL Id : hal-00153303 , version 1

Citer

Christine Sinoquet, Sylvain Demey, Frédérique Braun. A large-scale analysis for significance assessment of frequencies relative to potentially strong sigma 70 promoters - comparison of 32 prokaryotic genomes -. 2007. ⟨hal-00153303v1⟩

Collections

LINA-COMBI
297 Consultations
293 Téléchargements

Partager

Gmail Facebook X LinkedIn More