Re: [Rdkit-discuss] Definition of HBA differs from pipeline pilot

[Hongbin Y. <yan...@16...>]

Hi, Chris,
? ? Thank you very much for the suggestion. But I tend to tell my fellows to use?Lipinski' HBA in PP :).
? ? (BTW,?http://www.macinchem.org?is pretty good. I like the website and thanks for the "advertisement")
Cheers,




Hongbin Yang?
?From:?Chris SwainDate:?2017-06-21?14:08To:?rdkit-discussSubject:?Re: [Rdkit-discuss] Definition of HBA differs from pipeline pilotHi,
Many applications have multiple definitions of HBA/D from simple heteroatom counts to sophisticated SMARTS definitions, as long as they are documented and referenced I’d vote for keeping all definitions. It certainly helps if you want to go back and try to repeat published work.?
Cheers,
Chris

Dr Chris Swain BA MA (Cantab) PhD?CChem FRSC
Macs in Chemistry
sw...@ma...
http://www.macinchem.org





Message: 2
Date: Tue, 20 Jun 2017 23:52:48 +0800
From: "Hongbin Yang" <yan...@16...>
To: rdkit-discuss <rdk...@li...>
Subject: [Rdkit-discuss] Definition of HBA differs from pipeline pilot
Message-ID: <201...@16...>
Content-Type: text/plain; charset="gb2312"







Hi, Rdkiters,
? ?The definition of HBA in rdkit is (by Lipinski) :
32 ?HAcceptorSmarts = Chem.MolFromSmarts('[$([O,S;H1;v2]-[!$(*=[O,N,P,S])]),' + 
33 ??????????????????????????????????????'$([O,S;H0;v2]),$([O,S;-]),$([N;v3;!$(N-*=!@[O,N,P,S])]),' + 
34 ??????????????????????????????????????'$([nH0,o,s;+0])]') 

? ? But in pipeline pilot (3.5), there are two HBA definitions, one of which is Lipinsk's. I guess the other is the "first edition", which is defined as:22 ?# HAcceptor ?'[$([!#6;+0]);!$([F,Cl,Br,I]); 
23 ?# ????????????!$([o,s,nX3]);!$([Nv5,Pv5,Sv4,Sv6])]' 
? ?Does it mean that we should use the newest edition of HBA and?get rid of the default definition in pipeline pilot. These may change the?datasets filtered by rules such RO5.
(I am not sure whether the HBA defined in PP is the same as defined in Line 22-23. I made a test.?Abacavir have 7 (current edition) and 6 (old) respectively. And in PP, it also returned these two results).

reference:?http://www.rdkit.org/docs/api/rdkit.Chem.Lipinski-pysrc.html#NumHAcceptors?

Hongbin Yang ???

Research: Toxicophore and Chemoinformatics
Pharmaceutical Science, School of Pharmacy

East China University of Science and Technology?