From chemistry-request %-% at %-% server.ccl.net Fri Mar 1 09:56:28 2002 Received: from rwcrmhc53.attbi.com ([204.127.198.39]) by server.ccl.net (8.11.6/8.11.0) with ESMTP id g21EuS518975 for ; Fri, 1 Mar 2002 09:56:28 -0500 Received: from daveg.cachesoftware.com ([12.224.212.200]) by rwcrmhc53.attbi.com (InterMail vM.4.01.03.27 201-229-121-127-20010626) with ESMTP id <20020301145551.XHRR2951.rwcrmhc53.attbi.com -8 at 8- daveg.cachesoftware.com>; Fri, 1 Mar 2002 14:55:51 +0000 Message-Id: <5.1.0.14.0.20020301062721.02c5cb90[ AT ]pop3.norton.antivirus> X-Sender: (Unverified) X-Mailer: QUALCOMM Windows Eudora Version 5.1 Date: Fri, 01 Mar 2002 07:00:09 -0800 To: "Julien MICHEL" From: David Gallagher Subject: Re: CCL:Linear scaling algorithms for QM calculations Cc: In-Reply-To: <008d01c1c0b9$8511bee0$3506c3d4 {*at*} sungam> Mime-Version: 1.0 Content-Type: multipart/alternative; boundary="=====================_128207419==_.ALT" --=====================_128207419==_.ALT Content-Type: text/plain; charset="us-ascii"; format=flowed One recent publication suggests that the frontier orbitals on the hydrated enzyme (calculated by MOPAC 2000) correlate to the active sites. See: Effects of Hydration on the Electronic Structure of an Enzyme: Implication for Catalytic Function, Kazuki Ohno et al, J. Am. Chem. Soc. 2001, 123, 8161-8162 A MOPAC 2002 SCF on Crambin (641 atom protein) takes less than 3 minutes on a 1,000 MHz Windows PC and uses only 60 MB of memory. MOPAC 2002 with the MOZYME linear scaling algorithm handles up to 20,000 atoms, for full geometry optimizations and reaction coordinates, and includes the COSMO solvent field. The algorithm also scales approximately linearly in memory usage. MOPAC 2002 is available stand-alone or as part of CAChe software suite. As the method is fairly new, there are not many publications on it's application to proteins around just yet. However, there is more information at www.cachesoftware.com David Gallagher, Fujitsu At 01:36 AM 3/1/2002 +0100, Julien MICHEL wrote: >Dear CCLers, > >I recently read a few words about algorithms for pure quantum mechanical >calculations that scales linearly with the number of electrons in the system >studied. >I was surprised to read this as I thought such rate was not feasible for any >kind >of quantum calculation. I would be curious to read more about these >algorithms and how they work. Furthermore, I'd like to know if these >algorithms have been used to study fairly big molecular systems such as >proteins. I would be glad if someone could point some references or even >better internet resources focusing on that topic > >Julien MICHEL --=====================_128207419==_.ALT Content-Type: text/html; charset="us-ascii" One recent publication suggests that the frontier orbitals on the hydrated enzyme (calculated by MOPAC 2000) correlate to the active sites. See:

Effects of Hydration on the Electronic Structure of an Enzyme: Implication for Catalytic Function,  Kazuki Ohno et al, J. Am. Chem. Soc. 2001, 123, 8161-8162

A MOPAC 2002 SCF on Crambin (641 atom protein) takes less than 3 minutes on a 1,000 MHz Windows PC and uses only 60 MB of memory.

MOPAC 2002 with the MOZYME linear scaling algorithm handles up to 20,000 atoms,  for full geometry optimizations and reaction coordinates, and includes the COSMO solvent field. The algorithm also scales approximately  linearly in memory usage. MOPAC 2002 is available stand-alone or as part of CAChe software suite. As the method is fairly new, there are not many publications on it's application to proteins around just yet. However, there is more information at www.cachesoftware.com

David Gallagher, Fujitsu


At 01:36 AM 3/1/2002 +0100, Julien MICHEL wrote:
Dear CCLers,

I  recently read a few words about algorithms for pure quantum mechanical
calculations that scales linearly with the number of electrons in the system
studied.
I was surprised to read this as I thought such rate was not feasible for any
kind
of quantum calculation. I would be curious to read more about these
algorithms and how they work. Furthermore, I’d like to know if these
algorithms have been used to study fairly big molecular systems such as
proteins. I would be glad if someone could point some references or even
better internet resources focusing on that topic

Julien MICHEL
--=====================_128207419==_.ALT--