[Zinc-fans] Getting single molecule files and applying functional group filters

[John J. Irwin]

Hi Lena

Thanks for your email and your interest in ZINC.
> First, once i download and unzip subset files that i want, is there an easy way 
> to break the large mol2 file onto the separate molecule files?
>   
You can use molchunk.py (a script supplied with OEChem). You can use 
splitmol (Daniel Gschwend '94), which comes with UCSF DOCK. It is also a 
nice short perl or python scripting exercise. If you can't write one 
now, I assure it is invaluable to learn how to write these little kinds 
of short scripts.
> Another question is more about docking but maybe you could help me. From 
> literature i understand that once a database is selected, one applies 
> physicochemical filters such as logP, MW etc.. to filter certain compounds out - 
> and ZINC allows one to do that.
>   
The property-filtered subsets such as lead-like and fragment-like apply 
physicochemical filters.
> After that, however, there are also filters for undesirable functional groups 
> (30 to 60 depending on citation) that are applied. As far as I understand, there 
> is no such filtering in ZINC. And if so, I imagine that there have to be some 
> kind of applications that would allow one to do that in an automated way. Could 
> you let me know what program(s) i could use to do that?
>   
I suggest you try OpenEye's filter program for removing undesirable 
compounds. It is both easy to use and, like my son's Pokemon, incredibly 
powerful.

ZINC does apply functional group filters. The problem is that 
"undesirable", like beauty, is in the eye of the beholder, and indeed 
also may depend on the project context, target, assay, etc. We filter 
out molecules that we think are problematic, either for docking, or in 
typical assays. Our views on this have changed, and continue to change 
(look for two papers from our group that address this topic in 2007).


Good luck!

John