COMPUTER AIDED TWO-DIMENSIONAL ELECTROPHORETIC GEL ANALYSIS

GELLAB-II软件系统是一个探索性的数据分析系统

The GELLAB-II software system is an exploratory data analysis system for the analysis of sets of 2D electrophoretic protein gel images. It incorporates sophisticated subsystems for image acquisition, processing, database manipulation, graphics and statistical analysis. It has been applied to a variety of experimental systems in which quantitative and qualitative changes, in one or more proteins among hundreds or thousands of unaltered proteins, is the basic analytic problem. Keeping track of changes detected using these methods is also a major attribute of the system. A composite gel database may be "viewed" under different exploratory data analysis conditions, and statistical differences and subtle patterns elucidated from "slices" of an effective 3D database. Results can be presented in a variety of tables, plots or derived images and on workstations over wide area networks. The GELLAB-II research has resulted in a technology transfer CRADA with CSPI/Scanalytics Inc., for a commercially available system GELLAB-II+, making this technology easily available to cancer researchers on inexpensive Microsoft Windows PCs (released in 1996). Such commercialization results in wider use and better support of the GELLAB-II technology than we can provide. GELLAB- II applications this period include: ongoing studies of nuclear matrix protein changes in prostate cancer to aid screening and staging of men with prostate cancer (J.Hopkins); and extending an earlier Rett syndrome study (CDC) Robinson MK, Myrick J, etal. Electrophoresis 16: 1176-83, 1995. In the prostate cancer studies,we are looking for both missing proteins as well as subtle quantitative changes in patterns on sets of proteins which correlates with experimental conditions and staging and screening of men with prostate cancer. We are also doing a similar investigation of nuclear matrix protein patterns from renal cell carcinoma and other genitourinary tissues. The Hopkins group's prostate molecular work has identified several of the protein spots by sequence analysis and are completing studies to confirm these sequences. GRABT=Z01BC08382 A complete understanding of the function of RNA molecules requires a knowledge of the higher order structures as well as the characteristics of their primary sequence. The two and three-dimensional structure of RNA are important for many functions, including regulation of transcription and translation, catalysis, and the transport of proteins across membranes. The understanding of these functions is important for basic biology as well as for bioassays and the development of drugs that can intervene in cases where pathological functionality of these molecules occur. We have been involved in developing computational approaches for improving RNA folding prediction and the analysis of the results. We have been developing and improving a novel RNA folding technique that uses concepts from genetic algorithms. We were the first group to apply this approach to this problem. Our genetic algorithm runs on a massively parallel supercomputer with 16384 processors, and has recently been adapted to other massively parallel supercomputers, a CRAY/SGI T3E and ORIGIN 2000. We have seen improvements in results compared with more conventional folding algorithms and have incorporated new approaches which speed up convergence, specify stopping criteria, and allow the prediction of tertiary interactions (H-type pseudknots). Besides the GA we have also expanded our computational RNA analysis workbench, STRUCTURELAB, to include new analysis algorithms that permit the phylogenetic comparision of large databases of RNA structures. We also have the ability to generate 3-dimensional atomic models of various folded RNA structures. We have used these algorithms in studying various strains of HIV and the enteroviruses. This includes the 3-dimensional modeling of a portion of HIV.