Numerous restriction mapping programs have been devised since 1978, but none of them is as precise as manual map construction using digest gel photos, pencil, and graph paper. Given realistic data, programs usually find hundreds or even thousands of solutions, only one of which can be correct. Because of this, slow and demanding manual techniques are still in common use. The researchers propose to develop public domain software implementing a complete restriction mapping environment which will be far more powerful and useful than current restriction mapping software. This system will: 1) find solutions more quickly than any existing software; 2) find far fewer false maps than any existing software; 3) allow the user to """"""""steer"""""""" the entire mapping process (if desired); and 4) guide the user with detailed, expert advice on handling specific mapping problems as they arise. In addition, the system will have: 5) built-in extensibility, so that simple modifications will also allow other genetic marker mapping problems to be solved by the same program and 6) the ability to record all user activity transparently, providing quantitative data on successful mapping strategies. The system will be based on extant restriction mapping programs, but it will overcome their limitations by including the following additional capabilities: 1) All known heuristics will be implemented. (Heuristics are logical rules of thumb which guide the search towards a true solution.) Existing programs use only a small subset of the known heuristics. Every added heuristic will speed up the search and reduce the set of possible solutions to a problem. 2) Hand mapping will be simulated on-screen via a """"""""what you see is what you get"""""""" graphical user interface, with users choosing fully automatic mapping (the default), fully manual mapping (with simulated pencils and log paper), or various in-between levels of semi-automatic assisted mapping. This will allow users to control or adjust any part of the mapping process if they so desire. 3) An expert system (a program which can answer queries and make decisions by consulting a knowledge database) will guide users through the mapping process. It will assist the process of data acquisition, help the user solve difficulties, and tutor inexperienced users. Considerable preliminary design work has already taken place, so implementation can begin almost at once. Software will be developed and tested iteratively (""""""""rapid prototyping"""""""") to assure end-user satisfaction. To insure portability at the source code level, the two major modules of the system (fragment length derivation and mapping) will be coded in C++ using Boochs' object-oriented design methodology, and the user interface will be designed using a portable interface-building tool that works on a variety of computing platforms (including DOS machines, and the Macintosh). The expert system will be implemented using CLIPS, a portable C-based public domain expert system shell.
| Inglehart, J A; Nelson, P C; Zou, Y (1998) Mapper: an intelligent restriction mapping tool. Bioinformatics 14:101-11 |
