FastROCS: Difference between revisions

From DISI
Jump to navigation Jump to search
(Created page with "==Overview== FastROCS is an virtual screening tool using shape comparison for potential actives discovery. It is vastly faster improvement of [http://www.eyesopen.com/rocs | R...")
 
 
(21 intermediate revisions by the same user not shown)
Line 1: Line 1:
==Overview==
==Overview==
FastROCS is an virtual screening tool using shape comparison for potential actives discovery. It is vastly faster improvement of [http://www.eyesopen.com/rocs | ROCS]. It could automatically split the search into many parallel searches depending on how many GPU are there.  
FastROCS is an virtual screening tool using shape comparison for potential actives discovery. It is vastly improved of [http://www.eyesopen.com/rocs ROCS] in term of speed. It could automatically split the search into many parallel searches depending on how many GPU are there.
 
==Tutorials==
==Tutorials==
*(Admin-only) Setting up FastROCS server
 
   View [[FastROCS Setup]]
 
*Running searches and other stuff
===Install FastROCS===
  View [[FastROCS Usage]]
See [[Openeye python libraries]].
There is also an installation guide on OpenEye website for alternatives. [http://docs.eyesopen.com/toolkits/python/quickstart-python/install.html Click here]
 
===(Admin-only) Setting up FastROCS server===
'''This needs to take place in a GPU-enabled computer'''
   View [[FastROCS Server]]
 
===Running query on FastROCS server===
Any computer
====1. Set up virtual environment====
On bash shell
$ source /nfs/ex9/work/khtang/openeye/FastROCS/source_venv
 
To deactivate conda environment
$ conda deactivate
 
====2. Run search====
$ source /nfs/home/khtang/ex9/openeye/FastROCS/source_Shape
Convert smiles into sdf molecules. I wrote a script that converts each smiles into individual sdf files
$ python /nfs/ex9/work/khtang/openeye/FastROCS/scripts/smi2sdf.py [smiles]
 
$ ShapeDatabaseClient.py [-h] [--tversky] [--shapeOnly]
                              [--alternativeStarts {random,subrocs,inertialAtHeavyAtoms,inertialAtColorAtoms}]
                              server:port query results [nHits]
example: ShapeDatabaseClient.py epyc-a40.cluster.ucsf.bkslab.org:8080 ZINC000000002837.sdf result.sdf 3000
 
Send a query to a specified ShapeDatabaseServer and print the histogram of scores for the entire database of molecules
 
$ ShapeDatabaseClientHistogram.py <server:port> <query> <results> [num_hits = 100]
 
===Running FastROCS on command line on a small database===
 
 
[[Category:Tutorials]][[Category:Khanh]]

Latest revision as of 20:23, 7 October 2022

Overview

FastROCS is an virtual screening tool using shape comparison for potential actives discovery. It is vastly improved of ROCS in term of speed. It could automatically split the search into many parallel searches depending on how many GPU are there.

Tutorials

Install FastROCS

See Openeye python libraries. There is also an installation guide on OpenEye website for alternatives. Click here

(Admin-only) Setting up FastROCS server

This needs to take place in a GPU-enabled computer

 View FastROCS Server

Running query on FastROCS server

Any computer

1. Set up virtual environment

On bash shell

$ source /nfs/ex9/work/khtang/openeye/FastROCS/source_venv

To deactivate conda environment

$ conda deactivate

2. Run search

$ source /nfs/home/khtang/ex9/openeye/FastROCS/source_Shape

Convert smiles into sdf molecules. I wrote a script that converts each smiles into individual sdf files

$ python /nfs/ex9/work/khtang/openeye/FastROCS/scripts/smi2sdf.py [smiles]
$ ShapeDatabaseClient.py [-h] [--tversky] [--shapeOnly]
                             [--alternativeStarts {random,subrocs,inertialAtHeavyAtoms,inertialAtColorAtoms}]
                             server:port query results [nHits]
example: ShapeDatabaseClient.py epyc-a40.cluster.ucsf.bkslab.org:8080 ZINC000000002837.sdf result.sdf 3000

Send a query to a specified ShapeDatabaseServer and print the histogram of scores for the entire database of molecules

$ ShapeDatabaseClientHistogram.py <server:port> <query> <results> [num_hits = 100]

Running FastROCS on command line on a small database