| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: FightAIDS@Home
Thread: Evaluation of local optima |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 6
|
|
| Author |
|
|
steffen_moeller
Cruncher Joined: Dec 3, 2005 Post Count: 44 Status: Offline |
I understood that the docking energy presented is the mean over a sample of 200 or so mutations found in patients. If this is so and if this is the only information kept then I do consider this as problematic. What about a drug that is top performing on one or two thirds of the mutations but absolutely miserable on all others? I think we would want to learn about such, too, maybe for using it in combination with other partial solutions to achieve a full coverage. Please be so nice to educate us on your internal evaluation of our calculations.
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hi steffen_moeller,
Only a project scientist can answer your post with assurance, but - - I understood that the docking energy presented is the mean over a sample of 200 or so mutations found in patients. Where do you find any mention of 'mean'? Nothing I have read about FightAIDS@Home states this. Your argument against this method seems obviously correct. Therefore, I do not believe that we are using such a method. mycrofth |
|||
|
||||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
To clarify the issue of docking energy for each computation. Firstly, the docking energy that we compute is our estimate of the Gibbs free energy of binding between the ligand and the target protein (ie. the HIV protease ). This includes both enthalpic and entropic terms. The lower the predicted energy the better the binding. Any given computation running on your computer matches one ligand against one protease variant. We will accumulate these results and look for the best compounds (not just those that have the best “mean energy” over the entire panel of mutants). Certainly those compounds that are effective against a subset of mutants will be explored more fully. So do not worry that we will be throwing out any promising leads! I hope this answers your questions.
Thanks for your support, concern, and interest. Art Olson |
||
|
|
steffen_moeller
Cruncher Joined: Dec 3, 2005 Post Count: 44 Status: Offline |
I am very glad to hear that the accumulation of results is performed at the server side and no significant preselection is done by the client. I googled for my source but could not spot it either, sorry. Admittedly, I also thought "it cannot be", so I asked.
Many thanks for your kind reply. Steffen Möller |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hi Everyone,
I have updated the FAQ on the Scripps FightAIDS@Home web pages, adding a new question, "What are the units of energy?" and the answer, along with some useful links. Thank you for all your valuable idle cycles. Dr. Garrett M. Morris Staff Scientist Olson Laboratory, The Scripps Research Institute |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Good to see that the new project has also brought new scientific posts to the forum. It's always good to get this kind of feedback
 |
||
|
|
|