Hi all,
Any help regarding the issue we're facing is more than welcome.
We are deep sequencing a 96-bp PCR amplicon as follows:
Using NEB high-fidelity Q5, we did a first PCR of 20 cycles with tailed genome-specific adapters and then a second PCR of 10 cycles using IDT Stubby SI adapters. Cleaned up using 0.8X Agencourt AMPure beads and loaded onto a MiSeq V3 150bp.
We obtained ~600k reads per sample. Now, a bizarre observation: in all samples, we consistently find 1% of the reads (which is tenfold higher than other random errors) to contain one particular G-to-T SNP which really shouldn't be there. Now it gets even more bizarre: looking at only those 1% of reads that contain the SNP, they all have a significant q-score dip only and exactly at that SNP location (see attachment). The reads which support the WT allele don't have such a dip, or at least much less pronounced.
Does anyone have any idea what could be the cause of this 1-bp SNP/q-score dip?
Many thanks!
Any help regarding the issue we're facing is more than welcome.
We are deep sequencing a 96-bp PCR amplicon as follows:
Using NEB high-fidelity Q5, we did a first PCR of 20 cycles with tailed genome-specific adapters and then a second PCR of 10 cycles using IDT Stubby SI adapters. Cleaned up using 0.8X Agencourt AMPure beads and loaded onto a MiSeq V3 150bp.
We obtained ~600k reads per sample. Now, a bizarre observation: in all samples, we consistently find 1% of the reads (which is tenfold higher than other random errors) to contain one particular G-to-T SNP which really shouldn't be there. Now it gets even more bizarre: looking at only those 1% of reads that contain the SNP, they all have a significant q-score dip only and exactly at that SNP location (see attachment). The reads which support the WT allele don't have such a dip, or at least much less pronounced.
Does anyone have any idea what could be the cause of this 1-bp SNP/q-score dip?
Many thanks!