| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: FightAIDS@Home
Thread: Other Chemical Sets and quorum |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 3
|
|
| Author |
|
|
we45dfa35gh3476
Advanced Cruncher Joined: Apr 19, 2006 Post Count: 57 Status: Offline |
I hope that FightAIDS@Home expands beyond the NCI/ChemBridge sets (currently our largest set is 500,000) for more molecules (once we finish processing what we have). Grid.org uses 3.5 billion different molecules against cancer proteins. I realize that would take quite some time. Hopefully once we finish all the processing we can for ChemBridge/etc we can expand to some of the other chemical databases:
----------------------------------------http://www.chem.ox.ac.uk/cancer/molecules.html Other chemical databases include: "Asinex; ChemBridge, ComGenex, Contact Services (now known in the USA as ChemDiv Inc.), Maybridge , and Oxford Molecular's HTS Database" Are these publicly available sets, or would WCG have to purchase the information? Those sets (and other protein targets) would keep us crunching for a longggg time  Also - why don't we reduce the quorum for FAAH to 3? I believe FAAH uses a quorum of 5 currently, per a project update (unless this has changed): "we send the same candidate-drug/protease combination to 5 different computers." The WCG Help Defeat Cancer project uses a quorum of 3 (per the forums). We could process a lot faster if we reduced the quorum, I would think.  Jeff [Edit 2 times, last edit by we45dfa35gh3476 at Jul 21, 2006 9:59:09 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hello we45dfa35gh3476,
We are currently running with a quorum of 3. I just checked my Results Status page for my BOINC computer and a recent validated FAAH result showed that 3 work units were sent out on 13 July. One did not return so after 1 week, a fourth work unit was sent (to me on 20 July) which I returned later that day. When we first started running BOINC, we sent more than 3 work units at a time, but we went through a period of experimentation and changed that. TSRI acquires the chemical databases that we run on WCG. I doubt that we are about to run out. We need to acquire more processing power before this becomes a worry. Lawrence |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
3.5 billion is grossly misleading. That number is generated by including all the derived molecules. The real number is a mere 35 million. Also, there is no indication of the overlap between these sets. It is unlikely that each library is entirely unique.
FightAIDS@Home chose to start with the NCI diversity set because it can cover a representative sample of a far larger library. I understand the ChemBridge library can be prescreened and filtered to provide only suitable molecules. Remember, FightAIDS@Home needs to check each molecule against a large number of targets because there are so many mutated forms of HIV. This takes a lot more work, but Olsen Labs have come up with practical strategies to maximise the research done with as little computation as possible. Where we go from here will no doubt depend on the outcome of these current phases. |
||
|
|
|