Paper Details

Application of occam to biological sequence comparisons

Authors: Sturrock, Shane S., Salmon, Ian

Abstract:

One of the major uses of computers by molecular biologists is the alignment of protein or DNA sequences. Since the structure and function of proteins and DNA cannot be predicted directly from sequence data alone, the most useful procedure is to compare unknown sequences against a database of known sequences. Those with a high degree of homology are likely to have similar properties. At present, the databases are growing exponentially with doubling in around 20 months. Consequently, comparisons are taking proportionately longer to complete. We have applied various sequential comparative algorithms within a farming harness constructed in occam from a pipeline of transputers. The design of the farming harness for optimum link communication is discussed. The advantage our method offers is scalability; an important consideration as the databases multiply in content. The ability to tailor the algorithm to work on various size networks of transputers also allows the user to weigh time considerations against the total cost of the system. Our present implementation can perform around 600 comparisons per second between a query sequence of 50 residues and a database using 64 worker processors.

Proceedings:

Proceedings of WoTUG-14: Occam and the Transputer-Current Developments, Janet Edwards, 1991, pp 181 - 190 published by IOS Press, Amsterdam

This record in other formats: