rafe gutman

hey infidels, i have a question about sequence conservation. suppose i'm comparing a 3kb intron between about 30 species, separated by around 40-50 million years. i identify a few regions that are conserved among all 30 species (around 10bp per region). how do i demonstrate that this conservation is statistically significant? i can calculate a simple chi-squared value, with the "expected" value being (1/4)^N*3000, where N is the length of the region. does this value really mean anything though? i mean, shouldn't there be some similarity between these sequences even if they were functionless?

if anyone knows of a good method, or has any suggestions, i'd really appreciate it.

thanks

(p.s. this isn't for a homework assignment)

RufusAtticus

Try blasting your sequence and using the probabilities that come up from that.

You could also use the McDonald-Kreitman test to see if there is significant selection on the gene.