30431 Introduktion til Bioinformatik
David Ussery
Fredag, 12 November, 1999


A DNA Structural Atlas of Escherichia coli



Analysis of Chromatin-associated protein binding sites

1. IHF

The IHF sequence logo, based on an HMM alignment of 80 experimentally determined IHF binding sites.

IHF

Localisation of predicted IHF binding sites.

IHF binding sites


2. FIS

The FIS sequence logo, based on an HMM alignment of 101 experimentally determined FIS binding sites.

FIS



Localisation of predicted FIS binding sites.

FIS binding sites

3. H-NS

The binding site for H-NS is a bit more tricky, since there are few documented strong H-NS binding sites. However, H-NS binds to sequences which are curved (and usually AT-rich). Although we can't obtain a sequence logo and predict H-NS binding sites, we CAN look at the DNA in upstream of genes known to be regulated by H-NS (we used 39 sequences), and compare the curvature and AT content of these sequences to the genomic background. So now the question becomes - does H-NS bind preferentially to curves or merely AT-rich regions?


H-NS binding to curves
The H-NS binding sites are 5.1 std. deviations from the chromosomal average.

H-NS binding to curved-DNA regions




H-NS binding to AT-rich regions
The H-NS binding sites are 8.2 std. deviations from the chromosomal average.

H-NS binding to AT-rich regions





H-NS shows a strongest preference for AT rich DNA, and a weaker preference for curves.




Back to the TIGR conference "electronic poster" main page



Back Back to the Lecture notes


Back to the CBS homepage
Back to Dave's Courses page

Last modified Thursday, 11 November, 1999 by David Ussery