Research Project
2793676
DESCRIPTION (Applicant's abstract): We aim to design and produce a software product for DNA fragment assembly that meets the needs of large-scale sequencing efforts. Such efforts demand software that is fast and highly accurate, can assemble huge numbers of sequences, requires minimal coverage, and correctly handles repeated regions. DNASTAR's proposed product will meet these needs through the use of a new linear-time layout algorithm that can be extended to handle myriad fragment reads by the incorporation and improvement of highly accurate consensus-caller and through the use of statistical and algorithmic methods for handling repeats. The combination will form a unique package of the fastest, most accurate, and most flexible software available for DNA sequencing. We have implemented a prototype assembler that demonstrates the feasibility of our proposal. Our Phase I work will continue development by analyzing existing techniques for possible incorporation and by implementing novel ideas. An important component of the study is the comparison of our new assembler to commonly used methods with respect to accuracy and speed. During Phase II, we will continue our work by implementing solutions proposed in Phase I. In addition, we will address porting our new assembler to UNIX, parallelization and design and development of a user interface. In Phase III, the completed assembly software will be marketed to current and new customers in the desktop market as well as to those performing large-scale sequencing projects requiring UNIX software. PROPOSED COMMERCIAL APPLICATION: NOT AVAILABLE
