From jle&$at$&toyota.wavefun.com Thu Jan 19 13:37:56 1995 Received: from wavefun.com for jle%!at!%toyota.wavefun.com by www.ccl.net (8.6.9/930601.1506) id NAA01423; Thu, 19 Jan 1995 13:00:41 -0500 Received: from toyota.wavefun.com by wavefun.com via SMTP (931110.SGI/930416.SGI) for chemistry ":at:" ccl.net id AA02795; Thu, 19 Jan 95 10:00:37 -0800 Received: by toyota.wavefun.com (931110.SGI/930416.SGI) for %!at!%volvo.wavefun.com:chemistry%!at!%ccl.net id AA11701; Thu, 19 Jan 95 10:01:02 -0800 From: "Joe Leonard" Message-Id: <9501191001.ZM11699%!at!%toyota.wavefun.com> Date: Thu, 19 Jan 1995 10:01:02 -0800 Reply-To: jle ^%at%^ wavefun.com X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: chemistry /at\ccl.net Subject: Ghose-Crippen LogP estimation questions Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 Folks, I have been working with the Ghose-Crippen LogP estimation method (J. Comp. Chem. 9,80(1990)), and have several questions regarding the atom-typing rules described in the paper(s): 1. They refer to a program called CLASIF as being able to take structural information and generate the atom type. Obviously, having this program would "answer" all questions - is it available anywhere? 2. They make a distinction between "groups linked through carbon" (R-) and aliphatic groups (Al-). In this model, do aliphatic groups only contain carbon and hydrogen atoms, or can they contain any atom as long as they are singly-bound. Also, can H- match Al-? 3. They discuss aromatic groups - as with (2) are these carbon/hydrogen aromatics only, or can they contain heteroatoms? 4. Their Type 51 is called "H bound to alpha-C" - is this the alpha carbon of peptides or is it something else? 5. How are "formal oxidation numbers" handled for aromatic systems. The model refers to carbons with integer oxidation numbers. Since aromatic systems could have non-integer oxidation numbers (1.5, etc), should they be rounded up or down? In general, how are such systems handled? 6. Does Type 57 (phenol, enol, carboxyl OH) include Csp-OH oxygens? 7. Is Type 72 amide nitrogens RC(=O)-N or is RCO-N something like R-C-O-N compounds? 8. Is Type 76 R(=O)-NO2 or R-O-NO2? Looking at the other groups in Type 76, it seems like the former is correct... 9. The model does not include cationic nitrogens - has this been dealt with in a later paper? Are there later papers? 10. Type 78 is Ar-N=X and X-N=X - what about Al-N=X? 11. The Types associated with Sulfur seem incorrect. All specify R- groups, which I assumed to be -C-???, but there is at least one example (sulfapyridine) which has S(O2)-N- bonding... Should the sulfur groups be "any atom" connections, or am I misunderstanding the terminology throughout? I realize this is a large number of questions, but they're basically the same kinds of problems. I think folks have implemented this model in their work (and if Arup and/or Gordon are lurking, I'd appreciate their input!), so there should be some expertise out there - how folks have resolved these ambiguities. Thanks in advance - I welcome all comments! Please flame me directly, unless they're particularly interesting (I've always hated reading boring flames). Joe -- ------------------------------------------------------------------------ Joe Leonard Wavefunction Inc. 18401 Von Karman, Suite 370 Irvine, CA 92715 I am a professional... 714-955-2120 do not attempt this at home. 714-955-2118 fax jle[ AT ]wavefun.com