openmolecules.org Forum: Cheminformatics

Home » DataWarrior » Cheminformatics » TPSA (DW vs RDKit values differ)

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

TPSA [message #1025]

Mon, 03 August 2020 13:13

sansun
Messages: 50
Registered: April 2019

Member

I found disagreement between TPSA values calculated from DW and RDKit although both seem to follow same methodology by Peter Ertl et al. in J. Med. Chem. 43, 3714-3717 (2000).

In some case the difference is quite large e.g. 87 vs 252

However, Spearman correlation is > 0.9 for my data set (N > 20000).

Report message to a moderator

Re: TPSA [message #1028 is a reply to message #1025]

Tue, 04 August 2020 11:24

thomas
Messages: 747
Registered: June 2014

Senior Member

One reason may be that DataWarrior considers the largest fragment only by removing counter ions, water molecules and other small fragments from the main ingredient before the calculation.

Report message to a moderator

Re: TPSA [message #1029 is a reply to message #1028]

Tue, 04 August 2020 12:20

sansun
Messages: 50
Registered: April 2019

Member

I performed calculations after extracting the largest fragments and used the same file for both programs.
However, I find useful information on the following link which says RDkit doesn't use P and S atoms for TPSA calculation.

Does DW uses these atoms?

https://www.rdkit.org/docs/RDKit_Book.html

Report message to a moderator

Re: TPSA [message #1046 is a reply to message #1029]

Mon, 17 August 2020 14:01

thomas
Messages: 747
Registered: June 2014

Senior Member

DataWarrior sticks to the original publication, which includes contributions for S and P

Report message to a moderator

Previous Topic:	Basic Nitrogen
Next Topic:	Non-aromatic Het Rings

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Fri Aug 14 21:18:12 CEST 2026

Total time taken to generate the page: 0.00550 seconds