distinct tumor lane from blood lane

[markero]

Hi dear all,
I have a lot of whole genome sequenceing lanes from tumor and blood of cancer patients.
Now I want to know whether the tumor lanes are real tumor lane but not blood.
As you know, sometimes sample swap can happen due to many reasons.
Is there any good algorithm or tool (scripts) for that? or Any suggestions?
The mapping has been done.
Many thanks.

Markero

[henry.wood]

A quick way would be to do copy number analysis. I would recommend CNAnorm, mainly because I am one of the (minor) authors, but the wiki is full of others. You don't need to do a whole lane, just a random 1 million reads should be fine. Do different lanes using 'blood' lanes as control. Blood versus blood should be fairly flat. Most tumours have some kind of copy number damage, so should be reasonably lively.