SciLifeLab/TIDDIT

Inverted DUP

Closed this issue · 5 comments

dnil commented

A random thought: does the current clustering handle inverted dups? How are they classified?

Hello there!
Both yes and no! We do have all the information to classify inverted duplications, but TIDDIT is not really doing it.
Instead, such variant will be reported as an inversion, and the positions of this inversion will correspond to the duplicated region.
There will be no indication that this is an inverted duplication, but if you check the covM flag, it should be about 50% higher than the average coverage.
I should try to add some more information to this, sadly there is no "inverted duplication" tag specified by the vcf format, so we have to choose either to report it as an inversion or a duplication, or invent our own tag...Which would you prefer?

dnil commented

Subtypes perhaps? Like DUP::INV DUP::TDUP DUP::IDUP. Next will then be aggregates with major and minor type, like: INV::DEL-INV-DEL!

dnil commented

Ping on this @J35P312... 😊

I got some ideas! I will have a look at it next week =P.

dnil commented

Looks like we can close this now! If it needs more testing, feel free to reopen. Nicely done anyway! 🥇