A WORKBENCH FOR LARGE-SCALE SEQUENCE HOMOLOGY ANALYSIS

When routinely analysing very long stretches of DNA sequences produced by genome sequencing projects, detailed analysis of database search r esults becomes exceedingly time consuming. To reduce the tedious brows ing of large quantities of protein similarities, two programs, MSPcrun ch and Blixem, were developed which assist in processing the results f rom the database search programs in the BLAST suite. MSPcrunch removes biased composition and redundant matches while keeping weak matches t hat are consistent with a lar ger gapped alignment. This makes BLAST s earching in practice more sensitive and reduces the risk of over looki ng distant similarities. Blixem is a multiple sequence alignment viewe r for X-windows which makes it significantly easier to scan and evalua te the matches ratified by MSPcrunch. In Blixem, matches to the transl ated DNA query sequence are simultaneously aligned in three frames. Al so, the distribution of matches over the whole DNA query is displayed. Examples of usage are drawn from 36 C.elegans cosmid clones totalling 1.2 megabases, to which these tools were applied.