short reads aligner for short reference database

[capricy]

Hello, there,

My illumina reads are 90-150bp; however, the database sequences are in only 100bp-2000bp.

I have been told that the current aligners are not very effective in picking up the mapped reads under such circumstances.

Could anyone provide some advice about what to do then?

Going back to blast or blat? Just feel they are slow

Thanks)

capricy

[GenoMax]

Would it not be more useful if you actually tried popular current aligners (bwa, bowtie1/2, bbmap, novoalign, you decide on the list) with a sampling of your reads to see what actually happens and post the results here? Then we would have data to see if "current aligners are not very effective in picking up the mapped reads under such circumstances" is true.

[Brian Bushnell]

When aligning to sequences that are shorter than read length, it's best to tell the aligner to use local rather than global (aka end-to-end) alignments.

[capricy]

If local alignment mode is invoked, are these RNAseq aligner better than blast (or mpiBLAST) in terms of speed and sensitivity?

[Brian Bushnell]

Speed, yes. Sensitivity, well, it depends on the settings, of course. But they are more sensitive at a given speed.

[capricy]

Thanks for the input!