Using Varscan without a matched normal

Question

[sugus]

Hi there,

I got an extremely incredible use of varscan.

You see, typically, we need a matched-normal sample as an input for varscan for somatic mutation calling.

However, if I used a matched-primary tumor as so-called normal baseline, and another matched-metastasis tumor sample, does it mean anything when I use this combination to call mutation?

Excuse me for this weird question, I just wondering if it means anything.

Many thanks advanced.

Answer 1

[Chris Miller]

Somatic calling is essentially just calling differences between two samples. If you use a tumor as your control and a metastasis as your case, then you'll call metastasis-specific mutations (but will miss any somatic mutations shared between the tumor and met).

[sugus]

Thank you, I just need to confirm this is worth trying.

[Charles Warden]

I think you will need to have several rounds of analysis and discussion for your projects.

So, having exposure to a certain amount of ideas is important, but you also need first-hand experience to be confident in your results. In other words, if you think something is worth trying, then you should proceed with that test and use that experience to guide future decisions :)

However, I am very grateful for the wonderful feedback that I have received on Biostars, and I hope this helps give you the confidence to design tests / experiments to help you understand phenomenon better (based upon criteria that you understand and feel comfortable with). Best of luck for your project!

[sugus]

Yeah, i am working on this. Thanks for this sweet encouragement :)