Foreground species exhibit distinct amino acid residues at sites under positive selection

[张东]

Dear developers,

Thank you for developing this excellent tool. To identify a set of positively selected sites common to all foreground branches, I used the branch-site model of codeml to detect such sites under positive selection in the foreground branches. However, the result indicated that the 185th site of this gene shows strong signals of positive selection, yet this site is not consistent across the foreground branches. Some was K, others E, and still others P, as shown in the figure with the foreground branches marked in white. Is this phenomenon normal?

[Sandra AC]

Hi there,

Thanks for your message! Under the BEB section in the screen output, the second column corresponds to the AA at the site the first column mentions. In your case, your output would indicate that the first sequence in your alignment has a "K" at position 185, but other sequences may have another AA. We provided some details on this output in Álvarez-Carretero et al. (2023); see a quote from the latter below:

```
In each line, the first column shows the site position (e.g., 10, 25, 108, and 123), which is followed by the amino acid at this site in the first sequence (this is for identification of the site in the sequence). The third column (Pr (w > 1)) shows the posterior probability for the site to be from the positive-selection class (i.e., with ω > 1). The last columns show the posterior mean of ω and the standard deviation in the ω distribution for the site.
```

Hope this helps answer your question!

All the best,
Sandy

[张东]

Thank you for your reply.  I've understood your interpretation of the results. However, I am slightly confused by the fact that a highly significant signal of positive selection was identified by the algorithm at site 185 of this protein, but in the alignment I observed that site 185 is highly variable, with amino acids K, A, D, E across the foreground branches?
Is this a normal feature in strongly positively selected sites?

[Janet Young]

hi, 

I might be misunderstanding your question, but seeing highly variable amino acids is totally expected at positively selected sites. That is exactly the pattern that the positive selection analysis is looking for - recurrent amino acid changes at a higher rate than you'd expect given the number of synonymous changes.  Does that make sense?

Janet

[Ao Qo]

Hi,

I’ve actually been encountering the same issue in my own analysis.  Thanks for the clarification.

I do have a follow-up question: if we don’t require amino acid changes along the foreground branches to be in the same direction (e.g., all M to S), how can we distinguish truly positively selected sites from those showing diverse changes (e.g., M to K, M to S, etc.) that might instead reflect relaxed selective constraint rather than directional positive selection?

I’d appreciate your insights on how such patterns are typically interpreted or handled.

Best regards,
Na Wan