Hey everyone. I have a bam file which I converted to sam (via samtools) and then used Picard's SamToFastq script and split the bam into two fastq files. I am now trying to run the de novo assembler Trinity on these files. I know they are strand specific, but I do not know their orientation. Are they forward-reverse orientation or reverse-forward? Which strand is the left and which is the right? Does anyone know how to determine this?
I guess it depends on the method used for generating strand specificity. For dUTP method, I believe its first-strand reverse (hence, reverse-forward). However, since you have mapped your files already, from the flag of every pair or just a couple pairs, you can check if the read is first in pair or second in pair and if the read is reverse oriented or not, isn't it?
By converting the reads out of .sam format, you have gotten rid of all the mapping information, like orientation.
If you are expecting every read of read 1 to be in one orientation, that's generally not true, unless the sample was prepped in a strand-specific way.
Either way, eyeball a portion of the .sam, and look at the binary flags to learn the orientation of some of your reads. You could split your .bam so that one half has all the forward reads, and one has all the reverse reads, if that's what you want.