CLC Genomics Workbench - exclude certain genes from reference sequence

[mrs-sir]

Hi all,

I'm having a little trouble with my analysis using CLC tools.
I'm working with NGS data from a gene panel including 11 genes. These reads should be mapped against the human genome and compared to SNP data from several databases (dbSNP, COSMIC, etc.). However, the mapping and SNP calling results must include only hits within the 11 genes in the panel.

My initial approach was to prepare a reference FASTA file only consisting of the 11 genes. But then the positions of the called variants do not correspond to the chromosomal positions of the SNP data from the databases.

In the CLC manual I found something about masking the reference sequence. I tried to convert my prepared FASTA file to a track using the Track Tools, but I couldn't select a FASTA file for the conversion.

So, how do I mask my reference sequence so that just a certain group of genes are considered in the mapping and the output is still based on the full reference (chromosomal positions)?
I'm open and grateful for other suggestions.

Thank you in advance for your help.

Best regards!

[jonathanjacobs]

Are you using GW6?

Tracks is not the way to do this - you should use a Read Mapping file; not a Track file.

This might be a hack - but you might try annotating the genome with a custom annotation called something like "Excluded Regions" - and apply this annotation to the entire genome EXCEPT for the 11 genomic regions. Then, when you do your read mapping from your NGS data, use the Excluded Regions annotation as a masking element - that way reads only mapping to your genes of interest will be included.

Of course - you run the risk of having reads mapped to your genes that could potentially map somewhere else in the genome much better; so I would consider upping the read scoring metrics quite a bit to compensate.

I may not understand exactly what the issue is - but maybe that helps a bit.

[CLC bio - Anja]

Hi mrs-sir
I asked the guys in support about your issue and they can definitely help. Just shoot them an email with your issue at [email protected] - if you still are struggling with it
Cheers,
Anja from CLC bio