Software for PCs

From: thompson.ga@pg.com Subject: Software for PCs Date: 9 Oct 1997 13:09:57 -0400 We are considering moving from a UNIX to an IBM PC for running NONMEM and would be interested in graphical and statistical software recommendations. Thanks in advance Gary A. Thompson, Ph.D. Clinical Pharmacology Procter and Gamble Pharmaceuticals 11450 Grooms Rd. Cincinnati, OH 45242 513-626-2058 thompson/ga@pg.com

From: Jim Ong <jong@belmont.com> Subject: Re: Software for PCs Date: 9 Oct 1997 19:26:25 -0400 Gary, You may want to consider the CrossGraphs graphical reporting system, marketed by my company, Belmont Research of Cambridge, Massachusetts. CrossGraphs automatically divides data into subsets by distinct values or ranges of values of user-specified partitioning variables, and presents each subset of data in its own graph. This lets you create multi-page graphical reports which contain rows and columns of graphs, one graph per patient or patient subgroup, without requiring any programming. CrossGraphs is particularly popular at clinical pharmacology organizations within pharmaceutical companies. For example, you could use CrossGraphs to create graph arrays containing many drug concentration vs. time graphs, one graph per patient. CrossGraphs was selected as the standard graphing subsystem of the Pharsight Trial Designer, marketed by Pharsight Corporation in Palo Alto, California. It is a computer-assisted trial design software system which helps companies design more effective clinical trials by incorporating PK/PD models and patient population models into the simulation of clinical trials. CrossGraphs runs on Windows, Solaris, and HP-UX, and is currently in beta test on the PowerPC Macintosh. You can also develop custom user interfaces to CrossGraphs, to streamline common tasks carried out by specific groups of users and make the capabilities of CrossGraphs available without requiring them to learn the full product. Additional information about CrossGraphs, including examples of its application to various phases of pharmaceutical R&D, is available on the Belmont Research web page: http://www.belmont.com. Jim Ong -- Jim Ong, Director of Product Marketing Belmont Research 1250 Bayhill Drive, Suite 101 San Bruno, CA 94066 +1 650 635 0302 fax: +1 650 635 0303 jong@belmont.com http://www.belmont.com

From: Mark Sale <msale@medlib.iaims.georgetown.edu> Subject: RE: Software for PCs Date: 9 Oct 1997 20:38:18 -0400 Sam, I'm surprised we haven't heard more defense of Splus for nonmem post processing. The scripting language is excellent (especially for C programmers) and the graphics are outstanding. There is a lot of functionallity for smoothing, individual plots etc. But, it isn't cheap. I on the other hand manage quite well with Excel for the most part, but I use Splus sometimes. I have a nice collection of Excel macros to make plots from NONMEM output. Mark Mark Sale M.D. Assistant Professor of Medicine Georgetown University msale@medlib.georgetown.edu 202-687-8242

From: Suneui Hong <sehong@blue.weeg.uiowa.edu> Subject: RE: Software for PCs (S-plus 4.0) Date: 10 Oct 1997 01:41:53 -0400 Recently, S-plus version 4.0 is available with nice GUI (graphical user interface) and powerful graphics. Until version 3.3, all statistical analysese and graphics could be done by commands which was easy for the C or C++ programmers because S is a subset language of C for statistics. However, Version 4.0 enables the users without any programming background to do almost everything. It provides a spreadsheed style data manipulation and menu driven statistical analysis and graphics are possible. Another thing I want to mention is the Help system, which is different from usual applications running on windows or windows NT. Its help system provides all manuals just like books.. It is very easy to find what I want to see. Also, all the instruction is easy to follow. The S-plus company(StatSci) argue that "S-plus is not simply a programming language, like C or Fortaran. It is much more: a full-function calculator, a statistical package, a graphics tool, and so on". It provides the interface to C++ and Fortran and it looks quite efficient because we can use statistics modules and graphics modules on S-plus. I have been using S-plus for just 3 months.. I'm actually a novice. But I am already able to do most of statistical analysis and graphings on S-plus. I may be too much complimentary on S-plus.. but I felt it's worthwhile to say like that... Thank you.. Sunny Hong A Ph.D. student in Pharmacokintiec/Pharmacodynamics (Also a MS student in Statistics) College of Pharmacy, University of Iowa E-mail : sehong@blue.weeg.uiowa.edu Tel) 319)335-9735 Fax) 319)353-5646

From: "Paul Laub" <plaub@draco.rm.fccc.edu> Subject: Re: Software for PCs Date: 10 Oct 1997 01:51:14 -0400 Concerning software for NONMEM data processing on a PC, I have found the following three programs useful: 1. NMSEE, a FORTRAN program written by Alison Boeckmann of the NONMEM Project Group for data reduction. It compiles and runs well in DOS. This free program extracts the essential details from the often voluminous NONMEM output and writes them in a form easy to read or readily imported into a spreadsheet. 2. Perl, which is free and well documented. I write perl scripts to automate both the running and processing of NONMEM, which I run in DOS. Perl means "practical report and extraction language" and is an interpreted batch programming language. Though primarily meant for UNIX and its variants, there are version for Windows 95/NT as well as a 32 bit version for MSDOS. I use the later. Check out both www.perl.com and www.perl.org. 3. JMP, a statistics program with an intuitive, easy to use graphical interface. Yes, S-Plus is more powerful, but JMP is vastly easier to use. It is by far the best product I have used for data visualization. I still use S-Plus, but only when I have to. Mac and Windows versions of JMP exist. Developed by the SAS Institute, its algorithms are robust and dependable. It can be purchased for one time fee of ca. $600 (or much less with academic discount). See www.sas.com. Paul B. Laub Paul (Sisyphus) B. Laub mathematical modeling of biomedical data Dept. of Pharmacology 328 West Bldg. Fox Chase Cancer Center 7701 Burholme Ave. Phila. PA 19111 USA p_laub@fccc.edu (215)728-4743 (voice) (215) 728-2741 (fax)

From: Niclas Jonsson <niclas.jonsson@biof.uu.se> Subject: Re: Software for PCs Date: 10 Oct 1997 05:08:28 -0400 OK, here is some more S-PLUS stuff! I totally agree with Mark about the merits of S-PLUS. The graphics *are* outstanding and especially the Trellis Graphics library lends itself to the visualization of population data. (As it seems from the other responses to this question, there are some similarities between the functionality in CrossGraph's graphics and the Trellis Graphics in S-PLUS.) But in a addition to the strong graphic capabilities S-PLUS has almost limitless possibilities when it comes to statistical modeling. The downside of S-PLUS 3.3 for windows is that it is command line driven and has a rather steep learning curve. Even in S-PLUS 4.0 it will be necessary to use the command line since the GUI doesn't cover all the functionality (it doesn't even come close). The good news is that there is a program called Xpose (written by me and Mats Karlsson) that automates a large majority of the things you might want to do in S-PLUS in conjunction with a population analysis. Xpose is free ad can be obtained from www.biof.uu.se/Xpose. There is also a beta version of the next release of Xpose (2.0) that can be obtained from ftp.biof.uu.se. Note! The beta release is not in any way tuned for S-PLUS 4.0 so there might be trouble if version 4 of S-PLUS is used. Niclas

From: Mark Sale <msale@medlib.iaims.georgetown.edu> Subject: RE: Software for PCs Date: 10 Oct 1997 09:28:13 -0400 Dear nmusers, I've had a lot of requests for the Excel macro(s) for plotting NONMEM output. I really hestiate to send out software that I'm not willing to support. However, I'm currently working with MGA software on a project to provide an interface for NONMEM to be used as a post-processor for their clinical trials simulation package. I could split out the NONMEM front end and post processor and distribute that, with some degree of support. This will include some combination of point and click and code editting for the front end and point and click basic graphics back end. Some of you have seen parts of the front end. If you're interested, please email me. I won't promise it before Christmas though, we have a new software release scheduled for Oct 31. (BTW, the NONMEM stuff won't be in that release, probably the next). Mark Mark Sale M.D. Assistant Professor of Medicine Georgetown University msale@medlib.georgetown.edu 202-687-8242 fax 687-6437