Advantages of split by chromosome?

Question

[kwanghoon]

Hi, I'm a baby at this field, and trying to do some tutorials like GATK best practice, CNV analysis, HLA-typing etc.

I have a question about files that are split by chromosome.

What advantages do they have?

And If I use them, is my command correct like below?

ex) bwa mem REF_Chr1.fa A_1.fastq A_2.fastq > A.sam

After many process, do I have to merge them at the end?

Thank you!

[Pierre Lindenbaum]

What advantages do they have?

parallelisation.

[Pierre Lindenbaum]

is my command correct like below?

ex) bwa mem REF_Chr1.fa A_1.fastq A_2.fastq > A.sam

No never map a bam on a contig only. you'll get false positives. You must use the whole reference genome.

[kwanghoon]

Then do I need to split after bam processing using whole reference?

Sorry I have no knowledge.

[Pierre Lindenbaum]

you don't have to _split_ the bam. For example, for calling variants, most tools are able to call a specific region/chromosome if the bam is indexed (=bai file)