PubMed Health⌕ Search

Biomedical subjects

S Brossette

Publications and source records attributed to S Brossette.

2 recordsLinked to original sources

Pathology information systems: data mining leads to knowledge discovery.

Information systems in pathology provide opportunities for pathologists and clinical laboratory scientists to impact both clinical care and modern research agendas. The paradigm shift in health care from individualized care to population-based and standardized delivery systems has created both of these opportunities. In research, pathology information systems can provide key databases for health services research and new informatics-based approaches to database research. The latter is characterized by utilization of pathology databases for data mining to discover new patterns that provide new knowledge. The multidisciplinary knowledge discovery and data mining program at the University of Alabama at Birmingham focuses on this health care application, which has the potential to make a major impact on health care research and delivery.

Artificial Intelligence↗

A program for selecting DNA fragments to detect mutations by denaturing gel electrophoresis methods.

A computer program was developed to automate the selection of DNA fragments for detecting mutations within a long DNA sequence by denaturing gel electrophoresis methods. The program, MELTSCAN, scans through a user specified DNA sequence calculating the melting behavior of overlapping DNA fragments covering the sequence. Melting characteristics of the fragments are analyzed to determine the best fragment for detecting mutations at each base pair position in the sequence. The calculation also determines the optimal fragment for detecting mutations within a user specified mutational hot spot region. The program is built around the statistical mechanical model of the DNA melting transition. The optimal fragment for a given position is selected using the criteria that its melting curve has at least two steps, the base pair position is in the fragment's lowest melting domain, and the melting domain has the smallest number of base pairs among fragments that meet the first two criteria. The program predicted fragments for detecting mutations in the cDNA and genomic DNA of the human p53 gene.

Algorithms↗