Commercial and clinical use of DNA sequencing for analysis of genetic variation is currently limited, in large part, by the relative low accuracy and lack of automation provided by commercial analysis software, particularly when analyzing heterozygous or heterogeneous clinical samples. Based on our observations that many of the artifacts which make such analysis difficult are both context dependent and highly reproducible, we propose to develop pattern matching software to: (i) improve the reliability and automation of base calling; (ii) enable more accurate quantitation in trace analysis, particularly important for interpreting results from DNAs mixtures; (iii) generate robust quality-control measures; and (iv) if needed, even tailor the analytics to specific sequences of clinical interest. Our method will be independent of any specific polymerase, chain-terminator, or device. Our specific aims include developing (i) software for collecting databases of trace patterns; (ii) a set of standard DNA templates to help generate these databases, and (iii) software to match these patterns within traces. A phase-I pilot study to assay mixed DNAs containing sequences associated with AZT resistance in HIV-1 is proposed, and in phase II we may examine mutations in BRCA1 or other cancer-related genes.Proposed commercial application:We have identified immediate interest in this technology for commercial R&D applications in (i) mutation analysis for cancer pharmacogenetics, (ii) infectious disease load and population monitoring, and (iii) polymorphism discovery for genetic mapping. Long- term, the prospects for diagnostic sequencing, and even non-gel based DNA analysis (e.g. SBH or microsequencing) will rely on having available this technology for effective mutation discovery and analysis.National Center for Human Genome Research (NCHGR)
