How does SpliceMap work?
SpliceMap is a pure implementation of the split-reads alignment algorithm. 25bp half-reads are mapped to the genome with with either Bowtie, Eland or SeqMap. Then the half-reads are pieced together to determine locations of exons and possible junctions. Unmapped half-reads are used to search for splice junctions. Currently, only the cannoical GT-AG splice sites are identified.
In the case of reads longer than 50bp. The reads it separated into overlapping segments of 50bp and the same algorithm above is run. Finally, all the segments are stitched together to find the final alignment.
SpliceMap is geared towards mammalian genomes. If you are having trouble running SpliceMap on plant, yeast or bacterial genomes please let us know.