htseq-count

[moon_ligth]

Hello All,

I have some .bam files from samples of human, two group of sample: First one for Normal and the other for patient, I have no access for the reads files. and i want to generated the genes counts to use it with in edgeR. i am new to these things. Could any one please help me with the steps.
i directly tied the following command after i install the htseq-count

htseq-count -f .bam .gtf > counts.txt

from where i can get the the right .gtf file to use and do i need to put all the .bam files of every group together to get the counts.


Please help

[blancha]

Start by checking the header of one of the BAM files.
samtools view -H name.bam

At the end of the header section, check the data field CL (command line) in the PG (Program) line.
This will provide you with useful information about the GTF file to use, e.g. Ensembl or UCSC.

[blancha]

For your second question, no you shouldn't merge the BAM files together if each BAM file corresponds to one sample.

You will count the reads for each BAM file separately.
DESeq2 will take as input the htseq-count for each file.

Incidentally, if you are not able to identify the exact GTF file used from the header, you could always extract the reads from the BAM file and repeat the alignments. This would be more work however, and not absolutely necessary.