[nilshomer]

I like the presentation (heatmap table). Could you try varying the "-K" and "-M" parameters? Alternatively, you could design indexes with greater key size (more ones in the mask). There is a lot of flexibility, though I haven't thought about longer read lengths. The only criticism I have is that short-read aligners are designed for short-reads, and you are using them in a non-standard way; just like BLAT would not be used for 35bp error-prone reads.

[aleferna]

Yes actually the point of the study is to determine how they behave with different read lengths. Most mappers are made for say 50bp, but what about 55 or 45 or 60...? The problem is that my type of reads ( from 4C / Hi-C / 3C ) don't have nice single size reads. That's why I'm testing each aligner to see where it is the strongest and where it is the weakest. My hope is that I can combine 2 or more programs to have a complete solution.
My question for BFast is because I thought that a longer read would generate more CALS and therefore, all things being equal, would be easier to map ... which is what happens in Blat... anyway I will try to mess around with BFast a bit more