You can use following tools for protein pocket/cavity detection: PocketFinder, fpocket, or MetaPocket. Hope they'll help you

Hey Anita, did you try these software ? What did you think of it ? Easy to install ? easy to use ? which one is best according to you ?

Pocket finder is easy and online available, just input your PDB file, you don't even have to install it.

you can also use Active Site Prediction tool (http://www.scfbio-iitd.res.in/dock/ActiveSite.jsp) developed by IIT Delhi

and about the reliability of the predicted active site, you can check and confirm it through literature search.

you Q-site finder and Active site finder http://www.scfbio-iitd.res.in/dock/ActiveSite.jsp according to Anita Tripathi is right but according to your ligand you want to find the pockets then you can prefer the auto vina for analyzing the binding pockets

Hello Benjamin,

Never depend on any one server as all of them are based on different approaches. Normally i compare results of Castp & pocketFinder. Both of two servers give you idea of pockets i.e. of probable enzymatic sites and not of exact enzymatic sites!!! For exact enzymatic site try PDBsum or CAS (on EBI server), you can get results on these servers if enzymatic site has been reported by any scientist (they dont follow any bioinformatics approach to predict, if know they will tell you).

So best approach is to compare results of Castp, PocketFinder, PDBSum & CAS to conclude best pocket (which can turn into enzymatic site!!!).

And if you are working with Castp& PocketFinder then you don't need to run on any other servers could be Qsitefinder, Fpocket etc.

Also if you are using these sites for docking, then always use more then one site for docking you dont know where you ligand can bind in much better way!! Its more or less like hit-n-trial method.

All the best!

Hi Benjamin, we have recently developed a protein pocket detector using convolutional neural networks (DeepSite, link to paper here: Article DeepSite: Protein binding site predictor using 3D-convolutio...

Hi Ben, you can compare a few reliable programs as professional people mentioned above, while CASTp is definitely the good one. Now we updated CASTp to the new 3.0 release with more valuable information, and more user-friendly also. Please check with our server at: http://sts.bioe.uic.edu/castp

Benjamin B.L. Schwarz

And thank Amit Kumar Singh for your suggestions!!

Hi,

Active site prediction/ Protein pocket identification can be done using the following tools:-

i) 3D LigandSite

ii) COACH

iii) COFACTOR

iv) Active Site Prediction Server

All these tools are easy to use and the results are emailed to you rapidly.

Best regards,

Hello ,

You can use http://tcoffee.crg.cat/ to simulate the shape of your protein and detect the active site.