Hi Elaine,<br><br>Thank you very much for your detailed reply.<br><br>Nancy<br><br><br><br><div class="gmail_quote">On Sun, Jan 17, 2010 at 8:49 PM, Elaine Meng <span dir="ltr"><<a href="mailto:meng@cgl.ucsf.edu">meng@cgl.ucsf.edu</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">Hi Nancy,<br>
This is a general modeling question, not a Chimera question (since Chimera doesn't do automated docking) and would be better sent to a general forum such as <a href="http://ccl.net" target="_blank">ccl.net</a><br>
<br>
However, I'll attempt a brief and necessarily general answer.<br>
<br>
(a) It could be that the two binding sites are slightly different. This would not necessarily represent errors in the coordinates; perhaps the crystallization environment introduces slight asymmetries, or perhaps the two binding sites have some positive or negative cooperativity.<br>
<br>
(b) Even if the binding sites are "really" symmetrical, very small differences in coordinates, well within experimental uncertainty even at high resolutions, could cause large differences in the results. Say one atom is shifted by 0.001 A -- that could cross some cutoff in the method and prevent the ligand from fitting into one copy of the site.<br>
<br>
(c) Even if the coordinates for the two binding sites are exactly the same, there may be asymmetries introduced by how the docking program describes the sites or how it docks ligands into the sites. For specifics, you would have to look at the documentation for that program, or contact its support address, if any.<br>
<br>
I hope this helps,<div class="im"><br>
Elaine<br>
-----<br>
Elaine C. Meng, Ph.D.<br>
UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab<br>
Department of Pharmaceutical Chemistry<br>
University of California, San Francisco<br>
<br></div><div><div></div><div class="h5">
On Jan 17, 2010, at 5:14 PM, Nancy wrote:<br>
<br>
<blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
Hi All,<br>
I am performing molecular docking simulations of a ligand binding to a homodimeric protein, to determine a potential binding site(s). Due to the symmetrical nature of a homodimer, I would expect that the binding site(s) on one protomer would be identical on the other protomer. Therefore, a ligand should bind with equal probability and affinity to both sides of the protein. However, when I perform a molecular docking simulation (using an X-Ray crystal structure), the ligand preferentially binds to one side of the homodimer.<br>
<br>
Is this outcome likely the result of errors inherent in the X-Ray crystal structure, as one would expect identical binding to both sides?<br>
Thank you very much.<br>
Nancy<br>
<br>
</blockquote>
</div></div></blockquote></div><br>