Ultra-Conservation In Genome Comparisons

Question

[Haibao Tang]

Part of my research interest is to look for unusual conservation patterns in pairwise genome comparisons. Naturally, such patterns might suggest functional elements, or occasionally, lateral DNA transmission between organisms.

I have so far relied on ad-hoc rules to extract these. However, the same rule does not apply to a different species pair - for example, some would find a 200bp perfect match between chicken-human unusual, but it is a lot less surprising for chimp-human comparison.

My question here is whether there is a statistical method to measure the "surprise" of an alignment, given the expected sequence divergence between two species? I realize there could be complications as selective pressure varies for different types of sequences - but would like to see what others approach this problem.

Answer 1

[Khader Shameer]

Have you looked at the extensive genome comparison studies reported by Inna Dubchuk's group?

I would recommend you to start with the manuscripts that described genome comparison based on multiple alignments and pairwise alignments from the group (dataset, methods). Pre-computed data and an associated comparative genome browser Vista could be useful for you.

Answer 2

[Aaron Statham]

http://ancora.genereg.net/ and http://www.ncbi.nlm.nih.gov/pubmed/18279518 may be useful background

[Haibao Tang]

thanks for the pointer. According to the paper, "We identify HCNEs by scanning pairwise BLASTZ net whole-genome alignments (nets) downloaded from the UCSC Genome Browser database for regions with at least I identities over C alignment columns." - this is an arbitrary rule that I mentioned, which I hope that there is a better alternative to.