The primary objective of this phase I Contract is to port MLAB to run in a suitable UNIX environment, and convert MLAB to operate as a """"""""servr"""""""". The MLAB server will receive MLAB commands in the form of test strings, and return its results as text strings containing scalars, matrices, or text. For demonstration purposes, the MLAB server will be accessed via a """"""""client"""""""" program simulating PROPHET II which accepts user input from the keyboard, passes the input as a string to the server, using interprocess communication, and displays the resultant text strings obtained from the MLAB server, using interprocess communication. In Phase II simplified interfaces will be developed for use by biochemical researchers, who typically do not have the mathematical skills to utilize the current version of MLAB. The MLAB mathematical modeling program is unsurpassed in its ability to define mathematical models and fit them to data. PROPHET II is limited by comparison to MLAB, especially in three areas: definition of models based upon a rich collection of arbitrarily-defined composite functions, robust handling of mathematical exceptions, and use of symbolic partial derivatives in computing the Hessian matrix for curve fitting. By making the entire MLAB program available to PROPHET II, the limitations can be overcome. In addition, MLAB offers hundreds of built-in functions in the areas of data analysis, applied mathematics, and statistics not currently available in PROPHET II which will then be accessible to PROPHET II users.
