Reader Comments

Post a new comment on this article

Isn't this invalidating the method?

Posted by raik on 26 Jun 2010 at 12:48 GMT

Surprisingly, the similarity P-value analysis of these 53 pairs revealed that the physico-chemical similarity of the mimicking binding sites is not significant. However, their structural similarity is obvious, see Figure 2.
http://ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1000821#article1.body1.sec3.p5

The cases of obvious structural mimicry are a perfect positive control for the method used in this paper. The fact that what we know is similar is not actually classified as similar puts, in my opinion, a big question mark on this work.

No competing interests declared.

RE: Isn't this invalidating the method?

juliettemartin replied to raik on 28 Jun 2010 at 08:43 GMT

Dear Raik

In this paper, I counted the number of similar atoms/residues at the interface among all the atoms/residues that are structurally aligned. By similar, I meant atoms or residues from the same physico-chemical class.

I would not say that obvious cases of structural mimicry are a perfect positive control, since the similarity between them is structural, i.e., the shape of the binding sites. Shape is not taken into account in the present study.

Juliette Martin

No competing interests declared.