openmolecules.org Forum: Functionality » Generating evolutionary library

Home » DataWarrior » Functionality » Generating evolutionary library (Generating evolutionary library of 1M compounds)

Show: Today's Messages :: Polls :: Message Navigator

Generating evolutionary library [message #2241]

Wed, 17 July 2024 11:12

sublimeuser
Messages: 6
Registered: June 2021

Junior Member

Hello, everyone!

Would like to ask your advice on the following aspect: my aim is to generate a evolutionary library based on a starting compounds where at the end I will have roughly 1 million resulting compounds.

I want to use different fitness criteria (logP, logS and so on) while possibly protecting a certain portion of the parent compound that is known to be involved in binding.

Any suggestions on how I can convince DW to generate a more or less strict number of compounds using the Build Evolutionary Library functionality?

Many thanks!

sublime

Report message to a moderator

Re: Generating evolutionary library [message #2246 is a reply to message #2241]

Mon, 22 July 2024 16:45

thomas
Messages: 715
Registered: June 2014

Senior Member

1 million compounds is a lot. If you run 1000 generations, let 20 compounds every generation, and do that 50 times (50 runs) then you will get about one million compounds, provided that your criteria are not very strict and allow that many different compounds in related generations.

Typically, with strict criteria, e.g. docking, 3D-superpositioning, or a chemical similarity, you should reach reasonably optimized compounds with much less compounds. If you aim for diverse, druglike compounds, creating random compounds may be the better solution...

Report message to a moderator

Previous Topic:	Comparing two databases by merging
Next Topic:	Embedding credentials

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Sat Nov 23 09:20:37 CET 2024

Total time taken to generate the page: 0.03704 seconds