Figure 1

Strategy of JANE. 1) HSP regions (high scoring pairs) are collected applying the BLAST algorithm. Parameters were optimized to detect alignments with lower sequence similarity. 2) HSP fragments are processed using JANE's algorithm, a quality filter is applied to increase accuracy. 3) Top scoring HSPs for each reads are used as anchors. Next, further HSPs are selectively considered (only the closest first) if not too distant (criterion: remaining length of the EST/sequence read to be mapped). 4) The algorithm assembles and calculates potential coverage. Two reads sharing overlapping regions are consecutively merged, forming a contig. 5) A predicted genome is generated replacing more and more parts from the template genome by mapped and assembled sequences. 6) Modules in the JANE toolkit predict function information on individual ESTs or sequence reads.