The frontline support team will be unavailable to answer questions on April 15th and 17th 2019. We will be back soon after. Thank you for your patience and we apologize for any inconvenience!

FilterByOrientationBias in GATK4 hasn't filtered out any artefacts in FFPE sample


I am currently using GATK4 to identify somatic mutations from FFPE WGS data. Everything else works fine apart from the second filtering step trough FilterByOrientationBias. Since the given sample is FFPE, I assume there should be some artifacts (C>T) being filtered out in this step. However, it doesn't filter any.

The command lines I used:

java -jar $picard CollectSequencingArtifactMetrics I=tumour.bam O=tumour_artifact FILE_EXTENSION=.txt R=$genome

gatk FilterByOrientationBias -AM 'G/T' -AM 'C/T' -V "$sample"_somatic_oncefiltered.vcf.gz -P tumour_artifact.pre_adapter_detail_metrics.txt -O "$sample"_twicefiltered.vcf.gz

Part of the summary of the filtering:

Sample Artifact_Mode Num_Artifact_Mode Num_Artifact_Mode_Filtered
Tumour C>T 11175 0
Normal C>T 0 0
Tumour G>T 3494 0
Normal G>T 0 0

As you can see, there are over 10K artifacts from FFPE and 3K from OxoG. But none of them have been filtered. Is it because the tool is still under development or is it because I did something wrong?

Looking forward to hearing from you soon!

Best wishes,


Sign In or Register to comment.