From chemistry-request@www.ccl.net Fri Jan 22 02:48:55 1999 Received: from euler.central.cranfield.ac.uk (euler.central.cranfield.ac.uk [138.250.48.9]) by www.ccl.net (8.8.3/8.8.6/OSC/CCL 1.0) with ESMTP id CAA16130 Fri, 22 Jan 1999 02:48:54 -0500 (EST) Received: from btmodelling.pc.cranfield.ac.uk ([138.250.75.1] helo=btmodelling) by euler.central.cranfield.ac.uk with smtp (Exim 1.92 #2) for CHEMISTRY@www.ccl.net id 103bKn-000111-00; Fri, 22 Jan 1999 07:48:53 +0000 Message-Id: <3.0.3.32.19990122074527.00971b90@mailboxes.ccc.cranfield.ac.uk> X-Sender: bi0100@mailboxes.ccc.cranfield.ac.uk X-Mailer: QUALCOMM Windows Eudora Light Version 3.0.3 (32) Date: Fri, 22 Jan 1999 07:45:27 +0000 To: CHEMISTRY@www.ccl.net From: Richard M Day Subject: Leapfrog Summary Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Quite a while ago I sent the following message regarding the package LEAPFROG de novo design here is are some of the the replies: ORIGINAL EMAIL ! Dear All, I am trying to find people with experience in using LEAPFROG to suggest lead affinity peptides. The big question - does this work at all ?? What is is the best way of sifting out the definate non-running peptides before synthesis ? Is it worth seeing if one of the low energy conformations of the peptide is close to the one required to provide interaction with the binding site - i.e the one suggested by the lead finding program. Your Comments would be greatfuly appreciated, Thanks Richard D. ------------------------------------------------------------------ Hello Richard I've been looking at Leapfrog's peptide method recently as a reasonable way of coming up with readily synthesizable ligands for a crystal structure of an unligated receptor. >Dear All, > >I am trying to find people with experience in using LEAPFROG to suggest >lead affinity peptides. > >The big question - does this work at all ?? (A) Does it work in producing reasonable peptides which seem to fit well in to the site of interest? Yes and no. It has great potential, but in our hands we find the ionization states of Asp, Arg etc unacceptable and further, it does not cap the termini appropriately. Coincidentally, Steve Bowless from Tripos was out here yesterday and I talked to him about this and it seems he wants this fixed, as it should be easy to do and will greatly improve results. (B)Does the set of compounds it comes up with represent an improved set over random peptides? (i.e. do more of the compounds that leapfrog thinks will bind better actually bind better?) Because of the capping problem, the interaction energy estimate doesn't mean much. Even when this is fixed, it has to be realized that the binding estimate is extremely crude, so one would think about testing the top 100 peptides perhaps, rather than just the top 10. Nevertheless, by virtue of the fact basic steric and crude electrostatic matches are passed in this set, I would think it would be a better quality set than a random set. The problem is here, how many false negatives are we missing because crude treatment of receptor flexibility: no protein x-ray structure for which there is high res data is exactly the same at that with its complex (Murcko, Rev. Comp. Chem.)? One thing I want to look at is passing the Leapfrog set through a secondary scoring function, involving HINT (G.Kellogg), or VALIDATE (Marshall), as these really might be more meaningful. One problem I have about the Tripos fitting functions is that relative to Kuntz's DOCK, electrostatic interaction seem to be over-emphasized relative to a good geometric fit. I think of the fit as the primary thing, and electrostatics second, but if you have Arg and Asp, this interaction dominates in Leapfrog. I spoke to Steve Bowless about this and he's going to look into it. If it works, the best workaround will be to scale down the contribution of electrostatics by setting the dielectric to a high distance value, but we don't know yet if Leapfrog looks at what one has set for this. >What is is the best way of sifting out the definate non-running peptides >before synthesis ? > >Is it worth seeing if one of the low energy conformations of the peptide is >close to the one required to provide interaction with the binding site - >i.e the one suggested by the lead finding program. > > A worthwhile thing to do here would be to write an SPL macro to prioritize the hitset according to the number of "allowed" versus "disallowed" phi and psi angles etc. It is suprising, but as far as I know, there has been no study to look at the success of this sort of approach. It would be very interesting to screen a few thousand tripeptides and screen a 100 Leapfrog tripeptides and compare and publish the results. Or, take a known peptide ligand (e.g. RGD) and see if Leapfrog comes up with it if one has a reasonable starting site geometry. Jonathan Baell ---------------------------------------------------------------------- Dear Richard, My major is strcuture-based drug design and I know something about LEAPFROG. Your question is actually the most vital in the area of structure-based drug design: can we predict the activity before we do the experiment? Unfortunately the answer at present is "we cannot always do that". LEAPFROG is often criticized for its naive scoring method. My suggestion is: do not trust its ranking of the candidate molecules. But LEAPFROG is a good tool to build molecules within the binding pocket of the target. At least the molecules generated have relatively good shapes complementary to the binding pocket. You can choose yourself the molecules which seem to be plausible and make necessary modifications on the structure to enhance the interaction between the receptor and the ligand. Of course, it will depend on your own expertise. So you see why we always talk about "computer-aided drug design" rather than "computer-done drug design". Hope these help. Good luck. Arthur ---------------------------------------------------------------------- From chemistry-request@www.ccl.net Fri Jan 22 03:58:32 1999 Received: from mail.matav.hu (mail.matav.hu [145.236.224.244]) by www.ccl.net (8.8.3/8.8.6/OSC/CCL 1.0) with SMTP id DAA16352 Fri, 22 Jan 1999 03:58:27 -0500 (EST) Received: (qmail 21641 invoked from network); 22 Jan 1999 09:58:02 +0100 Received: from line-209-87.dial.matav.net (HELO inhale) (145.236.209.87) by mail.matav.hu with SMTP; 22 Jan 1999 09:58:02 +0100 Message-ID: <36A83D62.83CB0CB5@chemaxon.com> Date: Fri, 22 Jan 1999 09:57:06 +0100 From: Ferenc Csizmadia Organization: ChemAxon X-Mailer: Mozilla 4.01 [en] (WinNT; I) MIME-Version: 1.0 To: chemistry@www.ccl.net Subject: New ideas for chemistry software X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit A few weeks ago I asked your opinion about future trends in chemical software development. The first responses have already been published in CCL (ftp://www.ccl.net/pub/chemistry/archived-messages/99/01/10) The second batch follows: 11. ==== The summary you posted contained many interesting good ideas. I was disappointed, however, in that most of the answers focused on the needs of computational chemists rather than the needs of chemists in general. Typical answers of the former type were "development of hot new method" and "porting software to hot new machine". Computational chemistry can impact chemistry in two ways: 1) by making larger numbers of people into computational chemists, i.e., chemists who enjoy spending lots of time with a computer and never go into the lab, and 2) by providing "lab" chemists with useful information for their work. I think the latter is more important in the long run. Computational chemists and software developers certainly need to look at ways of making their software more powerful, more friendly, more portable, etc., but it won't count for much unless they also find ways to make computations more relevant. -Alan ----------------- Alan Shusterman Department of Chemistry Reed College www.reed.edu/~alan 12. ==== > It seems nearest future trends in chemistry are more or less visible. > To make it short: Biology, Biochemistry and related topics. > With regard to this discussion, I would like to add that the REASON why computational approaches and software are so enticing in biology, biochemistry, etc. is because of the difficulty in obtaining high-quality experimental data -- quickly and at low cost. The suggestion was made later in this discussion that if one wanted to start a new software project that one should look at structural biochemistry. I don't think that someone starting a new software project should necessarily limit themselves to this one area. There are many pressing needs in biology, biochemistry and molecular medicine OTHER than in structural modeling. There are many pressing needs in other areas of chemistry and engineering. Certainly, a software developer should also NOT neglect to examine the development of low-cost "user-friendly" software for data acquistion and control of analytical and chemical instrumentation. If profit and funding are the sole criteria for determining what software/computational project to start, then our accomplishments as scientists are doomed. I don't think that someone starting a new project should fall victim to the "bandwagon effect". From a business standpoint, it is difficult for a new software developer to break into a saturated market. It seems that profit and funding are the result of how we PROMOTE the results or application of the work to a potential grant agency, industrial user, etc. Our prime consideration in the work done should be how IMPORTANT is the problem solved or the application. Respectfully, Anita I. Zvaigzne 13. ==== Please keep in mind that PC does not mean Microsoft ! There have been several statements that want to get programs ported to Windows to be able to run them on "cheap" PC-hardware. Since there are many OSs that run these programs on exactly that cheap hardware, I would not see this urgent need for porting but would like the programs to be enhanced in functionality ! PS: I know at least 10 OSs of very different flavour running on exactly this PC hardware, of which MS_DOS(+Win3.1+Win9x) and Windows NT are only two, and will be only one really soon, since the DOS branch is declared death already by Microsoft. Moreover basically all of this can be installed in parallel on the same machine. So I would strongly suggest to work on the science part of the programs, not the operating system topics. Greetings, Jochen ----------------------------------------------------------------------- Jochen K"upper Heinrich-Heine-Universit"at D"usseldorf jochen@uni-duesseldorf.de Institut f"ur Physikalische Chemie I http://www-public.rz.uni-duesseldorf.de/~jochen 14. ==== Of particular interest to me, at any rate, would be developments in ab initio codes for handling "condensed phase" effects such as solvation in liquids and lattice effects in solids. For example, how is the potential energy surface for internal motion of a molecule/ion/radical affected by 1) H-bonding in solution, 2) trapping in e.g. a zeolite, 3) lattice constraints in a molecular solid? In principle, of course, all these systems can be handled completely at whatever level of calculation is appropriate, but it would be nice to have ab initio software with a selection of schemes which allow emphasis on the detailed electronic structure at the "focal point" of the calculation and on a lower level of detail (i.e. forces and dynamics) elsewhere. Regards, Rod Macrae ------------------------------------------------- Dr. R. M. Macrae Muon Science Laboratory Institute of Physical and Chemical Research (RIKEN) ============= Thank you for the contributions. Ferenc ~~~~~~ Ferenc Csizmadia, Ph.D. ChemAxon Ltd. Valyog u. 7, H-1032 Budapest, Hungary http://www.chemaxon.com T:+3620 9570988 mailto:fcsiz@chemaxon.com From chemistry-request@www.ccl.net Fri Jan 22 08:48:23 1999 Received: from mailbox.univie.ac.at (mailbox.univie.ac.at [131.130.1.26]) by www.ccl.net (8.8.3/8.8.6/OSC/CCL 1.0) with ESMTP id IAA17506 Fri, 22 Jan 1999 08:48:19 -0500 (EST) Received: from barolo.itc.univie.ac.at (barolo.itc.univie.ac.at [131.130.39.19]) by mailbox.univie.ac.at (8.8.4/8.8.4) with SMTP id OAA76944 for ; Fri, 22 Jan 1999 14:47:45 +0100 Message-ID: <36A88158.3E4C@univie.ac.at> Date: Fri, 22 Jan 1999 14:47:04 +0100 From: Werner Jakubetz Reply-To: werner.jakubetz@univie.ac.at Organization: Theoretical Chemistry/University of Vienna X-Mailer: Mozilla 3.0Gold (Win95; I) MIME-Version: 1.0 To: chemistry@www.ccl.net Subject: Last Circular: Winter-Workshop "Reaction Dynamics" Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit ___________________________________ WINTER-WORKSHOP "REACTION DYNAMICS" MARIAPFARR (AUSTRIA) ___________________________________ LAST CIRCULAR: The 1999 Theoretical Chemistry Winter-Workshop of the Austrian Chemical Society in Mariapfarr (a resort in the Austrian Alps) will take place from 16 to 19 February. Registration is still open until January 31. The 1999 workshop includes lectures on "Reaction Dynamics" and a mini-symposium open to all fields of theoretical chemistry. Lecturers: G.G. Balint-Kurti (Bristol): Time-Dependent Wavepacket Calculations in Reactive Scattering G.D. Billing (Copenhagen): A. Overview of Reactive Scattering B.  Mixed Quantum-Classical Dynamics R. Jaquet (Siegen): Quantum Reactive Scattering: The Time-independent Approach U. Manthe (Freiburg): Reaction Rates For details of the program, informations on the scope and format of the workshop and on Mariapfarr, and for all matters related to registration and accomodation, see the Mariapfarr-Homepage at http://www.kfunigraz.ac.at/tchwww/sax/mariapfarr/index_eng.html Werner Jakubetz _______________________________________________________________ Prof. Werner Jakubetz Institute for Theoretical Chemistry and Radiation Chemistry University of Vienna, Austria Address: Währingerstrasse 17 A-1090 Wien Austria Phone: +43-1-4277-52758 Fax: +43-1-4277-9527 email: werner.jakubetz@univie.ac.at _______________________________________________ From chemistry-request@www.ccl.net Sun Jan 24 15:49:32 1999 Received: from smtp4.fas.harvard.edu (root@smtp4.fas.harvard.edu [140.247.30.84]) by www.ccl.net (8.8.3/8.8.6/OSC/CCL 1.0) with ESMTP id PAA00435; Sun, 24 Jan 1999 15:49:31 -0500 (EST) Received: from login4.fas.harvard.edu (IDENT:daizadeh@login4.fas.harvard.edu [140.247.30.74]) by smtp4.fas.harvard.edu with ESMTP id PAA26799 Received: by login4.fas.harvard.edu with ESMTP id PAA25064 Date: Sun, 24 Jan 1999 15:49:27 -0500 (EST) From: Iraj Daizadeh To: chemistry@www.ccl.net, chemistry-request@www.ccl.net Subject: Smallest peptide with Structure (fwd) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hello. I was wondering what is the minimum number of amino acids required to construct a distinct seconday/tertiary structure in solution. I thought that there exist some brain hormones that are composed of no more that 15 or so amino acids. Your thoughts and particular references would be appreciated. Thanks in advance. Iraj. Iraj Daizadeh, Ph. D. Harvard University Department of Cellular and Molecular Biology The Biological Laboratories, Box #140 16 Divinity Avenue Cambridge, MA 02138 Phone: (617) 495-0783 (617) 495-0560 Fax: (617) 496-4313 Email: daizadeh@fas.harvard.edu