Service notice: Several of our team members are on vacation so service will be slow through at least July 13th, possibly longer depending on how much backlog accumulates during that time. This means that for a while it may take us more time than usual to answer your questions. Thank you for your patience.

Mutect2 Amplicon realigment behavior

EADGEADG KielMember

Hi there,

i was playing around with mutect2 and amplicons. I follow the best practise and run mutect2 with the following commands:

$java7Path\java -Xmx4g -jar $gatkPath\
-T MuTect2 \
-R $referencePath \
-I:tumor $I \
--dbsnp $dbSNP\
--cosmic $cosmic\
--artifact_detection_mode\
-bamout $bamOut\
-dt NONE \
-o $o

When i view the results a noticed that a evident call from the UG was missing in the mutect output. After i saw the bamOut from mutect, I release why... after the realignment all reads from the position were removed.

Here a view of both bam in igv:

Greetings EADG

Best Answer

Answers

  • Geraldine_VdAuweraGeraldine_VdAuwera Cambridge, MAMember, Administrator, Broadie

    I'm sorry, is this a question or just an observation? This is expected behavior, as documented for HaplotypeCaller. Realignment within the caller allows elucidation of such regions where the mapper makes mistakes.

  • EADGEADG KielMember

    A mashup of both ;) I was digging a little bit deeper into the HC-Doc..i assume the bottom of the picture "shows" the active region for my sample. Is the determination of the active region influenced by the number of Reads ? So that regions with higher read counts had a better chance to get marked as active region ?

  • Geraldine_VdAuweraGeraldine_VdAuwera Cambridge, MAMember, Administrator, Broadie

    Ah, got it. No, the active region triggering does not rely directly on the number of reads present; it's more a question of the proportion of reads showing non-reference bases. We don't want to be biased against low-coverage regions.

  • EADGEADG KielMember

    Sry for my late reponse, i was on vacation:)

    @Geraldine_VdAuwera said:
    Ah, got it. No, the active region triggering does not rely directly on the number of reads present; it's more a question of the proportion of reads showing non-reference bases.

    Ok, do I interpret correctly that the start of my reads are to "blurry" to be recognized as active region.

    Thanks a lot.

    Greetings EADG

  • EADGEADG KielMember

    Ty for the quick response...i used the "color by sample" option in IGV. I attach a screenshot showing the actual reads from the bamout.
    Can you anything more to it ;) ?

Sign In or Register to comment.