I wanted to see a particular sequence in our re-sequencing data using IGV. However the .bam file is too big (10Gb) to handle. Since I am interested in only a small region in the genome, I wondering if there is any way to generate a .sam/.bam file only for that region. Thank you very much in advance!
Chenglin
This has been answered many times before. You can specify a small region of your interest and create a small bam file as shown.
samtools view -bh chr1:100-200 > small.bam
•
written
by
Ashutosh Pandey
FYI, -b will always write the header if it's present (this isn't documented, but see here), so samtools view -b chr1:100-200 > small.bam will do the same thing.
Thanks Devon. Didn't know that.
•
written
by
Ashutosh Pandey
Thank you very much for your answer. Do I need to put the reference genome index file (My_genome.fasta.fai) in the command? Actually I typed something like this samtools view -bh chr1:100-200 My_genome.fasta.fai small.sam > small.bam, but I got error information [main_samview] fail to open "My_genome.fasta.fai" for reading. My index file is there, why can't it be read? Thanks!
You won't need the fai file at all.
The syntax is:
samtools view [OPTIONS] sorted_and_indexed.bam region
Edit: Just to make things clearer, in your case the command would be samtools view -b small.bam chr1:100-200 > small.subset.bam. The input must be a sorted and indexed BAM or CRAM file.
You can also use sambamba tool to slice the existing bam to specific regions. It is much faster than samtools.
Example : $sambamba slice -o my_output_bam_file.bam Input_bam_file.bam chr1:100-200
Or you can extract reads for an entire chromosome by just specifying the chromosome name in option
Example : $sambamba slice -o chr1_bam_file.bam Input_bam_file.bam chr1
samtools view -b in.bam chr1:100000-200000 > out.bam
did not work for me.
But
samtools view -b in.bam 1:100000-200000 > out.bam
worked. It depends on the genome used for the alignment. Using
samtools idxstats in.bam
can be useful to figure out how to name the chromosomes.
Log in to answer this question.
