Hey.
On the same paired-end sample I used (in Galaxy) TopHat (FR-Second strand) and Hisat2 (fr; Reverse). After that I used "Filter SAM or BAM, output SAM or BAM files on FLAG MAPQ RG LN or by region" function to cut a region of 20 kb and visualise it in IGV. The thing that puzzles me is that I have completely different results: in TopHat I have reads exactly in the region that I cut but in Hisat2 I have really long exons that exceed the region filtered by me with 30-40 kb and there are no reads. With Bowtie2 I did not manage to see anything in the same region. The reason that I am trying to replace TopHat is because I cannot process collections with it which is going to be really time-consuming. I used Hisat2 before and it worked just fine. But I do not know how to interpret the difference in the results between them. Any thoughts on how should I interpret the results? Thanks :)
Bowtie doesn't work because it is not a spliced aligner. So that makes perfect sense.
Right, but what about TopHat and Hisat2. Why Hisat2 gives me exons that are that long in comparison to TopHat when I introduce the same Library Type as an options (well, in this case the equivalent). I should add that I get long exons in TopHat as well but when I choose First Strand or Unstranded and not Second Strand. Could be because my library is not actually Second Strand as I thought initially? But Hisat2 gives me the same results for both Forward and Reverse.
It would be helpful if you showed the command-line used with the several aligners, and also (possibly maybe) show an example of some puzzling / discordant read mappings.
Got it. Just to make sure for future references because I am new here and I kind of training myself in bioinformatics talking with very few people about it (so I really have to learn the terminology better): I am using Galaxy. Should I specify the function used and the chosen option? I am not sure about the command line in Galaxy.