Distinguish sequencing errors and real mutations in mitochondria data

[cindylanzao]

Hi，all

We are searching a silico method to distinguish sequencing errors and real mutations from mitochondria sequencing data. Please do consider that the heteroplasmy in mitochondria. Thanks!

[Brian Bushnell]

Is this question in the context of high-throughput sequencing? Or do you just mean, in general, you want to study mitochondrial and you have no data yet?

[SNPsaurus]

SNP callers like Lo-Freq are meant for mixed populations. We used it when developing PELE-Seq (paired-end low error sequencing) which is a wet lab approach to detecting low frequency mutations. We examined cancer mitochondrial genotypes and were able to detect sub-population heteroplasmy. You could detect heteroplasmy with just Lo-Freq, but would need to discard SNPs below ~1-5% depending on the read depth and quality.

[cindylanzao]

Originally Posted by Brian Bushnell

Is this question in the context of high-throughput sequencing? Or do you just mean, in general, you want to study mitochondrial and you have no data yet?

Yes, NGS data

[cindylanzao]

Originally Posted by SNPsaurus

SNP callers like Lo-Freq are meant for mixed populations. We used it when developing PELE-Seq (paired-end low error sequencing) which is a wet lab approach to detecting low frequency mutations. We examined cancer mitochondrial genotypes and were able to detect sub-population heteroplasmy. You could detect heteroplasmy with just Lo-Freq, but would need to discard SNPs below ~1-5% depending on the read depth and quality.

Thanks for your nice method recommendation. Another problem would be how to measure the heteroplasmy level of mitochondrial DNA. Thanks!