Low tophat2 mapping of paired-end reads

[Aragion]

Hello. I'm trying to map paired-end reads on the reference scaffold using tophat2, but the percentage of mapped reads is almost 0%.

Left reads: Input: 9526882
Mapped: 224 ( 0.0% of input)
of these: 22 ( 9.8%) have multiple alignments (0 have >20)

Right reads: Input: 9526882
Mapped: 230 ( 0.0% of input)
of these: 18 ( 7.8%) have multiple alignments (0 have >20)

0.0% overall read mapping rate.
Aligned pairs: 111 of these: 5 ( 4.5%) have multiple alignments
0.0% concordant pair alignment rate.

I've tried to change tophat2 options (like intron length and inner distance) and to map unmapped reads on the same reference, but the result is always the same. Any suggestions are appreciated. Thanks.

[Brian Bushnell]

Some more information might be helpful here, like what your experiment is, what organism you're dealing with, sequencing platform, what the reference is, how you preprocessed the data, your command line, etc.

[GenoMax]

Any time question arises about low mapping percentage the first thing to do is to grab a few random reads from the files and do a quick blast search to make sure you are dealing with the data you think you have.