Question: Does high or low GC content affect base quality?
0
gravatar for Tania
13 months ago by
Tania120
Tania120 wrote:

Hi everyone

I understand that high or low GC content affect coverage, but could high or low GC content affect bases' quality too? And is there difficulty proving a variant by Sanger sequencing if it is in a high or low GC content?

Thanks

rna-seq • 620 views
ADD COMMENTlink modified 13 months ago • written 13 months ago by Tania120

I don't think there should be a big difficulty proving a variant by Sanger sequencing. Sounds like you found a variant that is not validating by Sanger sequencing?

ADD REPLYlink written 13 months ago by genomax69k

No, I have a variant with very low GC content. Found in all affected except one person. This person all reads agree with the reference but all bases are as low as 5 and mapping quality is also below 20. So was thinking does this low GC content could make terrible bases? Should I proceed to verify with sanger sequencing or this won't show up to if GC content is low. Note that we have many cases like this, all interesting genes, so if we discard all we might lose interesting thing.

ADD REPLYlink written 13 months ago by Tania120
1

Base qualities of below 5 and MAPQs below 20 don't sound good to me. 20 equates to a 1% chance of misalignment whilst 5 implies ~32% chance of an incorrect base call, if we interpret these scores literally.

High GC regions will have lower coverage because they relate to 'theremodynamically unfastened' regions that require more energy (heat) in order to separate the strands. If the strands cannot be separateed, they acn neither be amplified in cclonal amplification and cannot be sequenced. It has been shown that sonication coupled with a slightly higher melting temperature can help to even out coverage.

From such low quality data, I would regard Sanger confirmation as a must (?).

ADD REPLYlink modified 13 months ago • written 13 months ago by Kevin Blighe45k

Thanks Kevin a lot :) so my final point here does it worth Sanger sequencing? Like the variant exist in all affected patients, the low quality is in only one patient, lets say patient X. All other affected people have the variant with high mapping and quality. If we ignore the variant in patient x, we would say the variant is not segregating in all affected, then no need to Sanger sequencing as we should discard the variant completely if it is not segregating? do i make any sense? thanks a lot

ADD REPLYlink written 13 months ago by Tania120
2

Like the variant exist in all affected patients, the low quality is in only one patient, lets say patient X.

If this is the only candidate marker then by all means you should sequence by Sanger.

ADD REPLYlink written 13 months ago by genomax69k

great, thanks genomax. Was just checking it worth sanger sequencing not to simply throw it away :)

ADD REPLYlink written 13 months ago by Tania120
2

If there's a clinical decision resting on this variant call, then certainly it should be confirmed with Sanger sequencing, even if it is identified in all other patients. If your data is just research, you should probably still aim to re-do the sample as NGS, if possible. It really depends on what the significance of the finding could be for the patient and/or for your research group/company, and relates to how certain you want to be. In clinical settings, we cannot permit any doubt; in research settings, we can allow some amount of doubt to be included in our results.

ADD REPLYlink written 13 months ago by Kevin Blighe45k

Thanks Kevin, that's a nice explanation too :)

ADD REPLYlink written 13 months ago by Tania120
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 1698 users visited in the last hour