what happened to -rur option

[pemh...@gmail.com]

Hi there,

I am trying to find out the -rur option in the latest version. But it seems that it doesn't support it anymore. Any way to find out the unaligned reads? Thanks.

Peter

[Wan-Ping Lee]

Hi Peter,

MOSAIK currently keeps all reads in outpted bam including mapped and unmapped. So, you may use bamtools filter to catch them.

> bamtools filter -isMapped false -in <your_bam>

Best,

Wan-Ping

--
Wan-Ping Lee
Postdoctoral Research Associate
Department of Biology, Boston College

Tel: +1 (617) 552-2922
wanpi...@bc.edu

[pemh...@gmail.com]

Thanks Wan-Ping. I tried your suggestions and it works. I am now able to find the unaligned reads using both mosaik and bamtools.

Another question, do you know a way to split chimeric reads?

I read snownebula's reply on SEQanswer, but I don't know if it will fit our analysis. For example, he mentioned that one need to use a cut off for the minimal size of alignment such as 32 bp. In addition, one needs to use -mmal option. But I think -mmal option is not there anymore?

Any suggestions? Thanks a lot for your kind help.

Best,

Peter

[Wan-Ping Lee]

Hi Peter,

Could you please describe chimeric reads more? What kind of reads that you will think them chimeric?

Best,

Wan-Ping

[pemh...@gmail.com]

Hi Wan-Ping,

Thanks for the kind email reply.

In my work, we use NGS to sequence our samples and we find out that some reads map to different reference sequences. Our reference sequence contains three difference genes. So for a read map to more than one gene is consider a chimeric read as these three reference genes are completely different. I use blast to search for the match and half maps to one gene and another half maps to another gene.

As more than 10% of reads are chimeric, we'd like to split the chimeric reads up so that we could align them to their reference sequences well without causing too much variations (if it is not splitted) for SNP discovery.

Any suggestions? Thanks a lot for your kind help!

Best,

Peter

[Wan-Ping Lee]

Hi Peter,

I got your question. If the alignments are obtainable without splitting, I may try to modify cigar tags. I assume that those reads will be mapped with complex cigars, e.g. 20M1D1M3I2M5D2M. Then, I'll modify the cigar like 20M8S such SNP callers may not touch the complex part, 1D1M3I2M5D2M, and thus you may not hold too many false positive calls.

Best,

Wan-Ping

[pemh...@gmail.com]

Thanks Wan-Ping. I think I might have a way to get it. I am using bwa bwasw hard-clipping which will get rid of the unmatched region and give me the desired match portion. Thanks a lot!

Best,

Peter