From chemistry-request@www.ccl.net Sat Jan 23 18:16:01 1999 Received: from panix3.panix.com (hNnxrwAXgW851mVicUf6rmsMXBPXFIeJ@panix3.panix.com [166.84.1.68]) by www.ccl.net (8.8.3/8.8.6/OSC/CCL 1.0) with ESMTP id SAA26604 Sat, 23 Jan 1999 18:16:00 -0500 (EST) From: kynn@panix.com Received: (from kynn@localhost) by panix3.panix.com (8.8.5/8.8.8/PanixU1.4) id SAA05497; Sat, 23 Jan 1999 18:16:17 -0500 (EST) Date: Sat, 23 Jan 1999 18:16:17 -0500 (EST) Message-Id: <199901232316.SAA05497@panix3.panix.com> To: chemistry@www.ccl.net Subject: [Q] Histogram method I've come across a few papers on computational protein models that claim to use the so-called histogram method to estimate free energies of folding, but I don't see how they could have done this, except for cases much more trivial than those reported. It seems to me that to apply the histogram method, it would be necessary to run a simulation that sampled both the folded *and* the unfolded states reasonably well (so that the population histograms collected had any chance of approximating the real relative probabilities of the two macrostates). This means running the simulation at temperatures close to the folding transition temperature, Tf. But, except for very short chains (N < 75) or very streamlined models (cubic lattice models), folding/unfolding events become so rare as T approaches Tf, that it would take weeks or even months to collect the data for each application of this method. Is this so? I hope I'm wrong in this assessment, and that someone would be kind enough to clear up my confusion. Specifically, how is it possible to use the histogram method to estimate free energies of folding *without* collecting many folding/unfolding events near Tf? Thanks in advance, KJ From chemistry-request@www.ccl.net Sat Jan 23 23:13:56 1999 Received: from mail-03.telis.org (mail-03.telis.org [204.71.76.231]) by www.ccl.net (8.8.3/8.8.6/OSC/CCL 1.0) with ESMTP id XAA27664 Sat, 23 Jan 1999 23:13:54 -0500 (EST) Received: from [206.96.233.88] (s11-pm59.snaustel.campuscwix.net [206.96.233.88]) by mail-03.telis.org (8.9.0/8.8.8) with ESMTP id UAA20246 for ; Sat, 23 Jan 1999 20:08:49 -0800 (PST) X-Sender: gammadas@mail.telis.org Message-Id: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Sat, 23 Jan 1999 22:09:42 -0600 To: chemistry@www.ccl.net From: Goutam Das Subject: CPU comparison/G94W/G98W Hello CCLers! I wonder if anyone of you could please provide information on the relative performance of the AMD K6 2, Pentium II and Celeron processors when running G94W/G98W with ~96 MB RAM and 400 MHz and assuming every unit has 100MHz/66MHz bus and EIDE HD (or in other words, when everything else is the same). If anyone of you have had problems with the celeron and AMD K6 2 processors, when running these programs, I would like to know as well. If there are any reviews on this, please provide pointers. Thanx very much in advance Goutam Das GOUTAM DAS, Ph.D BETZDEARBORN (A division of Hercules) PO BOX 4300, 9669 GROGANS MILL ROAD THE WOODLANDS, TX 77387-4300 # 281.367.6201 xt 425///email:gammadas@telis.org