Hi, I am trying to use HISAT2 to align my RNA-seq data, but I am having some problems. I have 15 samples, all with stranded paired-end RNA-seq, and I am using GALAXY to process it. Now, I am in the "HISAT2" step to align it, but I think that something might be wrong, since some samples have a high percentage of reads with concordant alignment, but other samples have mostly discordant alignment.
As you see, there is a high variation in the alignment between samples, even though I have done exactly the same process for all of them (at this point, basically cutadapt with minimum length = 20 and Quality cutoff = 20).
I am using HISAT2 in Galaxy, configurating it for "Reverse paired-end" (fr), and I have not changed anything else from the standard configuration.
So, in conclusion. Is anything wrong with my process? How could I improve it? And, if not, is it normal that different samples have this variation in alignment?
Is this paired-end data? Did you happen to trim the data files independently? If you did you need to go back and trim the files in pairs.
I trimmed the data using cutadapt in pairs. Anyway, I did the same process for all samples, and some have great mapping whiile others is very discordant, so I don't get why is this so uneven.
Hello,
For RNA_Seq data, trimming is not essential because these analyses only certify the mapper reads and ignore all others.
Thks
Any possible solution? If not, is this discordant alignment problematic for further analysis?
Discordant alignments would be problematic for further analysis since you would not be able to get proper counts for affected samples. You should investigate what may be going on with these. Make sure the reads are in sync using a tool such as
repair.sh
from BBMap suite.