Aligning short reads to a human specific gene instead of reference genome

Question

[seta]

Hi all,

My focus of study is on the two human homologous genes related to blood groups, which have also hybrid forms due to gene-rearrangement. I'm going to detect various polymorphisms within them. I'm thinking of aligning short reads (PE, 100bp) to these reference genes from GenBank instead of the whole reference genome (hg19). Could you please tell me what's your opinion on this issue? If you agree with me, please kindly suggest me a suitable tool?

Thanks in advance

[GenoMax]

You could do that if the data is collected by a targeted method. If not you may be best off with aligning to the whole genome.

That said you could make a reference with the genes of interest and create indexes with NGS aligner of your choice.

[seta]

Thanks, the reads came from the whole genome sequencing. I read that aligning reads to the specific gene may be better resolve homologous genes, but I'm not sure about the correctness. Please kindly let me know what do you think?

Answer 1

[microfuge]

To add a bit more to @genomax ( " If not you may be best off with aligning to the whole genome."). The reference should be as complete as possible. In absence of a complete reference the aligner would force reads which come from the missing sequence in the reference to regions available in the incomplete reference. We had a case where the WGS data was aligned only to chloroplast genome and the results indicated a massive heteroplasmy with many tri/tetra alleleic SNPs. When aligned to complete reference these disappeared along with the hope of a big paper :).

[seta]

Thanks for your explanation... so, I go ahead with the reference genome.