[BioC] Rat Nimblegen promoter dataset HOWTO

Joern Toedling toedling at ebi.ac.uk
Thu Jan 17 18:05:05 CET 2008

I guess you will have to look at those differences between your two
groups in your data before thinking about appropriate follow-up
analysis. Are the differences rather of qualitative nature (enriched vs.
non-enriched) or quantitative (strongly enriched vs. weakly enriched)?
Besides, finding clear "enrichment" for histone modifications is
unfortunately really not that simple. Once you have regions enriched at
your chosen cutoff, however, there's a number of features that describe
these enriched regions, such as base-pair length, maximal fold-change
within this peak, area under the curve etc. There are lots of ways for
comparing these between two groups. Basic R packages provide many tools,
such as classical group tests, for solving these.


Dario Greco wrote:
> Dear Jim and Joern,
> thank you very much for your quick reply.
> I was just giving a look to the ACME and Ringo packages.
> as far as I understand, it is quite "simple" to look for enrichment of the 
> antibody over the input signals (Cy5/Cy3 in my case). 
> but what would be the best way to look for significant differences between the 
> two groups (treated vs control animals)?
> thank you once more,
> d

More information about the Bioconductor mailing list