Determining whether a mutation is beneficial, deleterious or neutral about dms_tools HOT 4 CLOSED

Yes. thanks.

from dms_tools.

For any given mutation, just take the ratio of the preferences. E.g., for site r if the wildtype residue is x and you want the effect of mutation to y, do \pi_{r,y} / \pi_{r,x} where \pi_{r,a} is the preference for residue a at position r. If this ratio is > 1, the mutation is favorable. If it is less than 1, the mutation is unfavorable.

By the way, I think the Stiffler and Ranganathan paper (http://www.ncbi.nlm.nih.gov/pubmed/25723163) has higher-quality lactamase data than the Firnberg et al paper, particularly if you take the Stiffler values for the highest antibiotic concentration.

from dms_tools.

I did as you said and the distribution of \pi_{r,y} / \pi_{r,x} looks like this.

\pi_{r,y} / \pi_{r,x} along the positions in the reference sequence,
In red, blue and gray are beneficial, deleterious mutations and wild type resp.

Yes, Stiffler et. al.'s data seems to be more comprehensive but unfortunately counts for synonymous mutations which I need are not available (even though their library is cloned with NNS primers strategy).

Thanks.

from dms_tools.

OK, these look like reasonably sensible results -- right?

from dms_tools.