what is difference from this project and one on world community grid
log in

Advanced search

Message boards : Science : what is difference from this project and one on world community grid

Author Message
kc0isw
Send message
Joined: 16 Jul 12
Posts: 1
Credit: 2,021
RAC: 0
Message 11 - Posted: 19 Jul 2012, 0:49:26 UTC

someone was going to ask it anyhow im sure

http://www.worldcommunitygrid.org/research/gfam/overview.do

Ant Chubb
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 18 Jul 12
Posts: 118
Credit: 1,019,875
RAC: 0
Message 14 - Posted: 19 Jul 2012, 11:03:29 UTC - in response to Message 11.

Hi,

The GoFightAgainstMalaria project led by Alex Perryman at Scripps aims to find new small molecule inhibitors of known protein targets. Their clever trick is to focus on mutations that have already been found in the established targets for antimalarial drugs. In effect it's humans against parasites - round 2, with humans adapting the fight to take into account the adaptations that the malarial parasite has evolved to resist the first salvo.
For this the GFAM project needs to dock millions of small molecules into a few receptor protein structures.

We're doing something very different. We're taking the small compound hits that have already been found to kill malaria, and trying to find which protein these compounds bind to. Big pharma have screened about 3 million compounds and found 19,000 that are active against malaria. But figuring out where these compounds bind in the parasite or how they work will be tricky. We hope to do this by docking each compound against each protein structure.

The main advantage we have over the GFAM project is that we KNOW that the compounds we're using already work against malaria. If we figure out which protein they inhibit, then that will open up whole fields of research - what does the protein do (biochemistry)? what does it look like (X-ray crystallography)? can we find better inhibitors (medicinal chemistry)? or is the one lead compound enough to rush through the clinic (clinical pharmacology)?

The main disadvantage is that we will never be able to cover the whole proteome. Some proteins are naturally disordered, and will remain opaque to crystallography or modelling. But we'll give it our best shot. We've already got structures (X-ray or models) for 1426 of the 5,363 proteins (26%), and we'll continue to do modelling of the more difficult proteins. If we were trying this with only one compound, we'd have a 1:4 chance of success. But as we're trying this nearly 19,000 times, we should have a fairly good chance of success.

I'd be delighted if we even only manage to get ONE new target in malaria, as this could be the Achilles heal that we'll be able to go for with the full might of medicinal chemistry.

New drugs developed from this open-science project will be against novel targets that the parasite has not had a chance to evolve resistance against. Like all drugs, they too will eventually fail, but at least we're trying to find new drugs faster than the parasite can evolve resistance. The fight continues...

I hope this answers your question.

ciao,
Ant

Profile ritterm
Avatar
Send message
Joined: 18 Jul 12
Posts: 13
Credit: 138,526
RAC: 0
Message 15 - Posted: 19 Jul 2012, 12:50:31 UTC - in response to Message 14.

Excellent overview, Ant. Thanks very much. Looking forward to crunching this project when the work starts flowing. :-)

Cheers,

MarkR

Ant Chubb
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 18 Jul 12
Posts: 118
Credit: 1,019,875
RAC: 0
Message 16 - Posted: 19 Jul 2012, 13:34:36 UTC - in response to Message 15.

Thanks for YOUR help. We can't do this without 1000's of PCs, so you guys are going to make all the difference.
:-)

ciao,
Ant

skgiven
Avatar
Send message
Joined: 27 Jul 12
Posts: 19
Credit: 490,044
RAC: 52
Message 71 - Posted: 27 Jul 2012, 20:36:14 UTC - in response to Message 16.

Some proteins are naturally disordered, and will remain opaque to crystallography or modelling. But we'll give it our best shot.

There are many other long established, and frequently overlooked, biochemical methods of identifying molecular shapes/structures, and of course NMR, which often better reflects the true functional structure and dynamic of a molecule than an X-Ray diffraction interpretation from a crystallized protein.
____________

Profile rilian
Avatar
Send message
Joined: 18 Jul 12
Posts: 1
Credit: 10,002
RAC: 0
Message 147 - Posted: 20 Aug 2012, 16:34:52 UTC - in response to Message 71.

thanks for the explanations!
____________

Profile Michael H.W. Weber
Send message
Joined: 5 Mar 15
Posts: 17
Credit: 303,100
RAC: 0
Message 2263 - Posted: 11 Apr 2015, 11:19:45 UTC - in response to Message 14.
Last modified: 11 Apr 2015, 11:22:54 UTC

The main advantage we have over the GFAM project is that we KNOW that the compounds we're using already work against malaria. If we figure out which protein they inhibit, then that will open up whole fields of research...

...yes, it certainly would, but:

The main disadvantage is that we will never be able to cover the whole proteome. Some proteins are naturally disordered, and will remain opaque to crystallography or modelling. But we'll give it our best shot. We've already got structures (X-ray or models) for 1426 of the 5,363 proteins (26%), and we'll continue to do modelling of the more difficult proteins.

Of these 1426 protein structures you mentioned, how many were experimentally determined (x-ray, NMR) and how many were (just) computer models (which makes docking difficult / unreliable)?
And regarding the computer models, who modeled these (your group and/or public database ressources?) and with which algorithms / software?
And of these models, how many are based on homology modeling and how many on de novo / ab initio modeling?

Michael.
____________

Profile Michael H.W. Weber
Send message
Joined: 5 Mar 15
Posts: 17
Credit: 303,100
RAC: 0
Message 2285 - Posted: 20 Apr 2015, 7:44:42 UTC - in response to Message 2263.

Any comments, please?

Michael.
____________

Anthony Chubb
Send message
Joined: 17 May 12
Posts: 9
Credit: 0
RAC: 0
Message 2287 - Posted: 22 Apr 2015, 2:36:54 UTC - in response to Message 2285.

Hi Michael,

141 proteins are X-Ray structures. 1534 are homology models. These came directly from the Sali lab who use Modeller. I used a cut-off score of 1.0 to select good models (as suggested by them - their score).

38 ligands are from co-crystal structures. 400 are the MMVmalaria box collection. The total of over 26,000 ligands is from the full GSK/Novartis collection and ChEMBL and liver stage hits.

Ciao,
Ant

Profile Michael H.W. Weber
Send message
Joined: 5 Mar 15
Posts: 17
Credit: 303,100
RAC: 0
Message 2288 - Posted: 22 Apr 2015, 16:21:35 UTC

Thank you for the feedback. ;-)

Michael.
____________

Message boards : Science : what is difference from this project and one on world community grid


Main page · Your account · Message boards


Copyright © 2017 Dr Anthony Chubb