GC Bias Correction

Question

[finswimmer]

Hello,

I'm doing DNA targeted resequencing. The target were enrichted by hybridization capture. In my bioinformatics pipeline I collect quality metrics wit CollectHsMetrics from picard tools.

From time to time there are some samples that have very different values for the GC_DROPOUT and AT_DROPOUT metrics compared to the other sample within the run. These samples made some problems in CNV analysis.

Is there a way to correct the coverage values for my target regions, taking the GC/AT-Bias into account?

fin swimmer

Answer 1

[ATpoint]

There is an implementation of the Benjamini & Speed method in deeptools, named computeGCBias, and a downstream tool, correctGCBias, that you might want to check out. Unfortunately I cannot tell you if this correction is necessary or beneficial on your dataset.

[finswimmer]

Thank you for your answer ATpoint.

I will have a look at this in the next few days. But it looks promising.

fin swimmer

[finswimmer]

Hello ATpoint,

a played a bit around with computeBias and it seems to make exactly what I need. If you move your comment to an answer I could mark it as accepted.

Thanks again.

fin swimmer

[ATpoint]

You're very welcome!