Forum:Discrepancy between Whole Exome Sequence result and Sanger Sequence result

Question

[haiying.kong]

I analyzed whole exome sequence data to identify somatic mutations in paired samples, 23 blood and 23 tumor samples from 23 patients.

Here is the list of software tools I used for the work. BWA for alignment, Picard for removing duplicates and sorting, GATK for realignment and base quality score recalibration, MuTect for somatic mutation identification, Oncotator for annotation. I used all default parameter settings.

For the same set of samples, my colleague did Sanger sequencing, and identified total 23 mutations.

My analysis on the whole exome data identified only 14 out of the 23 mutations that are identified from Sanger, and 2 additional mutations.

Could any one please tell me why the results from Sanger and Whole Exome can be so much different?

My colleague says the result from Sanger can be considered to be very accurate, and can be used to validate my result from whole exome data. If this is true, which parameter setting in which software tools could I tweak around to get better result?

Thank you so much in advance.

Answer 1

[Vivek]

You could do some debugging by checking if the mutations are not being called in either the tumor or blood or both to get a better idea of what's going wrong. If there are only 9 missing, you could possibly check the regions in BAM files using IGV. Sanger is indeed more reliable and an accepted method to validate mutations found from exome sequencing.