From chemistry-request@www.ccl.net Thu Aug 20 01:43:44 1998 Received: from jabez.anu.edu.au (jabez.anu.edu.au [150.203.193.193]) by www.ccl.net (8.8.3/8.8.6/OSC/CCL 1.0) with ESMTP id BAA09399 Thu, 20 Aug 1998 01:43:42 -0400 (EDT) Received: from localhost (really [127.0.0.1]) by jabez.anu.edu.au via in.smtpd with smtp (ident spg using rfc1413) id (Debian Smail3.2.0.101) for ; Thu, 20 Aug 1998 15:43:12 +1000 (EST) Date: Thu, 20 Aug 1998 15:43:10 +1000 (EST) From: Steve Greatbanks X-Sender: spg@jabez.anu.edu.au Reply-To: Steve Greatbanks To: chemistry@www.ccl.net Subject: DFT polarisation question Message-ID: X-No-Archive: yes MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hi folks, my questions relate to the applicability of DFT methods to the study of biological anion systems. I'll set the scene... I am interested in a biological molecule (folate) which has a formally doubly anionic group (a glutamate) at one end. Both HF and MP2 (using a range of basis sets, including diffuse functions) predict this to be the case, giving a charge very close to -2 (so far so good) for the GLU. My question relates to my DFT results (both LDA and GGA), which all give charges of ~-1.4 (rather less than I'd expect). Before I get picked up on this point, that is for Mulliken charges summed over the residue. The important point is that I get the same trend (i.e. DFT predicts the molecule to be far less polar than the HF or MP2) regardless of the charge partitioning scheme or basis set used. Does other people's experience agree with this? Are there any theoretical reasons why DFT should struggle to model the anionic system correctly? Interestingly, if I take the same molecule and embed it in a protein environment (using point charges), the molecule polarises substantially at the DFT level and I get the expected double anion, though the HF and MP2 results change very little (which is more what I'd expect). I'd be interested in hearing any opinions people have on this issue and pointers to any relevant literature I may have missed on the relative merits of HF and MP2 vs. DFT for anions, particularly their relative polarizability. I will, of course summarise any replies I get. Thanks, Steve Greatbanks -- +--------------------------------------+--+------------------------------+ |Dr Stephen P. Greatbanks | | | |Biochemistry & Molecular Biology |w:|http://biocomp.anu.edu.au/~spg| |John Curtin School of Medical Research|e:|Steve.Greatbanks@anu.edu.au | |Australian National University |p:|(+61) (02) 6279 8305 (JCSMR) | |ACT 0200, Australia | | | +--------------------------------------+--+------------------------------+ From chemistry-request@www.ccl.net Thu Aug 20 01:43:44 1998 Received: from jabez.anu.edu.au (jabez.anu.edu.au [150.203.193.193]) by www.ccl.net (8.8.3/8.8.6/OSC/CCL 1.0) with ESMTP id BAA09399 Thu, 20 Aug 1998 01:43:42 -0400 (EDT) Received: from localhost (really [127.0.0.1]) by jabez.anu.edu.au via in.smtpd with smtp (ident spg using rfc1413) id (Debian Smail3.2.0.101) for ; Thu, 20 Aug 1998 15:43:12 +1000 (EST) Date: Thu, 20 Aug 1998 15:43:10 +1000 (EST) From: Steve Greatbanks X-Sender: spg@jabez.anu.edu.au Reply-To: Steve Greatbanks To: chemistry@www.ccl.net Subject: DFT polarisation question Message-ID: X-No-Archive: yes MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hi folks, my questions relate to the applicability of DFT methods to the study of biological anion systems. I'll set the scene... I am interested in a biological molecule (folate) which has a formally doubly anionic group (a glutamate) at one end. Both HF and MP2 (using a range of basis sets, including diffuse functions) predict this to be the case, giving a charge very close to -2 (so far so good) for the GLU. My question relates to my DFT results (both LDA and GGA), which all give charges of ~-1.4 (rather less than I'd expect). Before I get picked up on this point, that is for Mulliken charges summed over the residue. The important point is that I get the same trend (i.e. DFT predicts the molecule to be far less polar than the HF or MP2) regardless of the charge partitioning scheme or basis set used. Does other people's experience agree with this? Are there any theoretical reasons why DFT should struggle to model the anionic system correctly? Interestingly, if I take the same molecule and embed it in a protein environment (using point charges), the molecule polarises substantially at the DFT level and I get the expected double anion, though the HF and MP2 results change very little (which is more what I'd expect). I'd be interested in hearing any opinions people have on this issue and pointers to any relevant literature I may have missed on the relative merits of HF and MP2 vs. DFT for anions, particularly their relative polarizability. I will, of course summarise any replies I get. Thanks, Steve Greatbanks -- +--------------------------------------+--+------------------------------+ |Dr Stephen P. Greatbanks | | | |Biochemistry & Molecular Biology |w:|http://biocomp.anu.edu.au/~spg| |John Curtin School of Medical Research|e:|Steve.Greatbanks@anu.edu.au | |Australian National University |p:|(+61) (02) 6279 8305 (JCSMR) | |ACT 0200, Australia | | | +--------------------------------------+--+------------------------------+ From chemistry-request@www.ccl.net Thu Aug 20 10:13:59 1998 Received: from s-cwis.unomaha.edu (s-cwis.unomaha.edu [137.48.1.11]) by www.ccl.net (8.8.3/8.8.6/OSC/CCL 1.0) with ESMTP id KAA13213 Thu, 20 Aug 1998 10:13:58 -0400 (EDT) Received: from cas.unomaha.edu (cas.unomaha.edu [137.48.34.40]) by s-cwis.unomaha.edu (8.8.5/8.8.5) with ESMTP id JAA08259 for ; Thu, 20 Aug 1998 09:09:24 -0500 (CDT) Received: from CAS/MAILQUEUE by cas.unomaha.edu (Mercury 1.21); 20 Aug 98 09:13:49 -600 Received: from MAILQUEUE by CAS (Mercury 1.21); 20 Aug 98 09:13:41 -600 Received: from zinc.unomaha.edu by cas.unomaha.edu (Mercury 1.21); 20 Aug 98 09:13:35 -600 Received: by zinc.unomaha.edu with Microsoft Mail id <01BDCC1B.6DF99B60@zinc.unomaha.edu>; Thu, 20 Aug 1998 09:18:01 -0500 Message-ID: <01BDCC1B.6DF99B60@zinc.unomaha.edu> From: "Douglas E. Stack" To: "'CCL'" Subject: X-Windows summary Date: Thu, 20 Aug 1998 09:18:00 -0500 MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by www.ccl.net id KAA13214 Dear CCLers, I was overwhelmed by the response to my original post which queried about an inexpensive X-windows program for some new PC in our student labs. The main purpose of the software is to connect to an SGI sever so that students can run Spartan remotely. There were too many replies for me to just paste the messages at the end of this note, so I will summarize the general theme expressed from this large response. 1. Exceed by Hummingbird, www.hummingbrid.com: Clearly the top choice if price is no option. What interest me is an available add-on the can run OpenGL. But the price is steep ca. $500, and more if the OpenGL option is purchased. Too much for student use, but I may get a copy for myself :). 2. PC-Xware by NCD, http://www.ncd.com/ppcx/ppcx.html: Also got high praise, second only to Exceed. The price is $395. They offer a 30-day demo, but I did not get a download for a test drive. 3. X-WinPro by Labf.com, http://www.ncd.com/ppcx/ppcx.html: A free demo version is available that runs for 30 min at a time. Did not download and try. Full version, ca. $100. 4. Exodus/Web TermX by Whitepine, www.wpine.com: I've used Exodus ver 5.6.4 for a couple of years on my Windows PC, it works great. It appears they've changed there product line with Exodus available for the Mac and Web TermX for the Windows platform. Our university has license agreement for Exdous (old version for both platforms) that cost ca. $120 per machine. This was the software is was going to buy because I've used it, it's simple, and has a lot of nice features. Read below how I've found a cheaper option. 5. X-Win32 by Startnet, www.starnet.com: This program does not have as many features as Exodus, but it does the essentials very well. This includes, logging on to remote severs or broadcasting through a local network for log-ons. A large selection of fonts, including some special add-on fonts; one of these was specific for SGI and worked great. Price: a fully functional free demo that runs for 2 hours before having to be re-started. I downloaded this demo version and try it for 2 hrs. Everything worked fine, and a warning message appeared 15 min before the two hour deadline. At that time, I log-off the sever, quit X-Win32, login again and was up for another 2 hrs. The non-demo version is $200, but the Web page describes site license agreements that could bring this price down. I think the demo option will be the route I take. The main purpose of this software is to connect to Spartan, students (and myself) very rarely will model for more than 2 hours at a time. If they do, the shutdown and restart takes about a minute and is not a real inconvenience. I've scan the Web page and carefully read all Readme files that came with the software, and I don't see any license limitation to using 6 copies of the demo for the purpose stated above. BUT, if anyone here, or more specifically, if any one from Starnet reading this, knows of any impropriety of doing this, contact me! Otherwise, I'm going to use the free demo for our students. Totally Free options: 1. MI/X by MicroImages, Inc., http://www.microimages.com/: The program was able to run Spartan remotely; however, it does not have any features for connecting to the remote server. One has to establish a connection with a telnet client first, type the command for the Xprogram while MI/X is running, and then the program appears in the single window provided. This is not as easy to use when compared to programs that provide an xdm login prompt and provide an SGI desktop. But since it total free (and since the company is located right here in Nebraska!), one should check this out for simple connections. 2. Linux by (insert your favorite Linux provider here): Of course many responses gave Linux as the best option for free X windows support. Don't get me wrong, I love Linux and I have a dual boot PC where I often turn to Linux when I need to have a lengthy logon to our server. But placing an additional OS on all six machines, and teaching students how to boot Linux and run the OS would take too much time for someone who is not yet tenured. While dual boots are handy, my next PC (the one I use in my office) will have just one OS on it. I still plan to run Linux, but on a dedicated PC. I've come to the conclusion of one OS for one PC. Thanks again for all the responses! Douglas E. Stack Assistant Professor Department of Chemistry University of Nebraska at Omaha Omaha, NE 68182-0109 (402) 554-3647 (402) 544-3888 (fax) destack@unomaha.edu