This software package contains two Windows programs, XLAGraph and Svedberg, developed at Amgen. The programs and other files are usually supplied via the Internet as a self-extracting archive, SVXLyymm.EXE, where yymm are the year and month (to indicate the versions). XLAGraph was originally developed to get around the problem of not being able to see XLA data as it is being acquired, especially during velocity runs. If you run your XLA data acquisition as a background task under Windows, then XLAGraph allows you to display the new data as it is generated. If you wish, it can automatically find the new files for all cells and graph them. XLAGraph is also useful for general graphing of XLA data and data transformations. It can read and overlay up to nine data files, with or without baselines subtracted. It is smart enough to load a whole sequence of scans and/or all cells for a scan in one operation. You can zoom in on any region of a graph by dragging the mouse, or manually set axes as you wish. XLAGraph can generate publication quality graphs on any Windows-supported printer, freeing you from the dreaded dot-matrix printer. (I don't even have one with my XLA.) It can also export publication quality graphs to other Windows programs. The usual data transformations to ln(c) vs. r^2 / 2 plots and of the x axis to Svedbergs are supported, again for up to 9 overlaid files. It can also write out new XLA data files which have had baselines subtracted and/or which are averages of other data files. XLAGraph is entirely menu and point-and-click driven. Svedberg is a data analysis program for velocity data. It simultaneously fits up to 9 absorbance scans to an approximate solution of the Lamm equation, to derive BOTH sedimentation and diffusion coefficients from the data. This means that you can also determine MW from a velocity experiment, and our experience is that this can be accurate within a few percent for proteins of 15-60 kDa. Svedberg is primarily intended for situations where diffusion makes the boundaries quite broad. It can also fit up to 3 species, under the assumption that they are entirely non-interacting, and has functions for both conventional and synthetic-boundary cells. A rigorous error analysis to find confidence intervals for the fitted parameters can be performed. The data sets may have a baseline data set subtracted, and/or an overall zero offset may be included in the fit. Svedberg is entirely menu and point-and-click driven. It uses the same graphics engine as XLAGraph to display high quality graphs of the data, overlaid data and fits, residual plots, etc., and all may be printed at publication quality or exported to word processors. Both programs were developed with Visual Basic. They will work with Windows running at VGA resolution (640 x 480), BUT they are really designed to run at 800 x 600 or higher. Windows 3.1 is required; 3.0 will not work! To successfully run XLA data acquisition as a background task you will probably need a 33 MHz 386 processor or better. The fitting functions in Svedberg are computationally intensive, and Visual Basic is not a true compiled language, so it wants a fast processor. On my 33 MHz 486 machine, a single species fit usually takes less than a minute, but I wouldn't want to do multispecies fitting on anything less. Documentation is almost non-existent, but others who have used these programs found them to be reasonably intuitive. Before using Svedberg you will want to read the article "Measuring Sedimentation, Diffusion, and Molecular Weights of Small Molecules by Direct Fitting of Sedimentation Velocity Concentration Profiles" from the "Modern Analytical Ultracentrifugation" book (Schuster and Laue, eds.) due out August '94, which describes the technique and gives examples. This article is available as a Microsoft Word for Windows 6.0 document, including figures, as SVEDBERG.DOC, from wherever you obtained the program files. It is also there as a Windows Write document, SVEDBERG.WRI, which contains most of the figures (but Write cannot handle some of them). I hope you find these programs useful! John Philo (jphilo@amgen.com) Protein Chemistry 14-2-A-223 Amgen Inc. Amgen Center Thousand Oaks, CA 91320-1789