Align to custom sequences using STAR aligner

Question

[roy.granit]

I have 3 amplicons that I have sequenced and now I wish to align them to the sequences of these three genes so I can quantify how many reads mapped to each gene.

In order to use STAR must I go through the steps of genome indexing or is there a way to align stright with the fasta file containing the 3 genes?

Thanks!

[lieven.sterck]

you could make (and index) a custom file with only those 3 genes (perhaps go for the gene locus sequences, rather than CDS or such). Keep in mind their is a risk that you bias your analysis.

Personally I think STAR is fast enough to use the whole genome as reference and do some filtering after mapping.

[roy.granit]

I did not mention that some of these genes do not exist in the genome..

[GenoMax]

If you don't expect off-target amplification then why not use whole genome indexes if they are available. That can also show you off-target amplification (if there was any).

If not, you will need to index the reference sequences (which should be trivial for a small number).

BBMap allows you to align directly without creating indexes first: bbmap.sh in=your.fq.gz ref=your_multifasta.fa out=aligned.bam

[roy.granit]

Thanks, how reliable is bbmap? never heard of it..

[GenoMax]

You are missing some important knowledge in that case :-) BBMap goes toe to toe with any NGS aligner out there. I will leave this link here so you can see what the BBTools suite can do (besides alignments). A guide for BBMap the aligner is available here.