Distribution of e-values in Best Reciprocal Blast Hits for Ortholog Mining

Question

[pld]

Does anyone have an idea of how the distribution of expect values will look when using reciprocal blast to detect orthologs?

Answer 1

[SES]

E-values are dependent on the length of the database, so they are not really comparable between analyses. Other reasons you can't directly compare the distribution of e-values is because the duplication history and divergence level between the species being compared might vary between analyses. I would look at the distribution of Ks values, that is commonly used and it will give you something easier to interpret.