CNV data in cBioportal and IGV

[jorgedela...@gmail.com]

Hi,

I am seeing that the CNV data in cBioportal does not look the same as in IGV. For example, a gene that is deleted in cBioportal for a specific set of data (is given a value of -2), then when you open the same data in IGV, sometime the gene is given a value  between -1 and 0. Could anyone explain me please why there's such a difference and how the CNV data is calculated in cBioportal? I would understand if a value of -1.5 is rounded up to -2, but from around 0.8 to -2 doesn't make sense to me. 

I would appreciate if someone could clear this up for me.

Many thanks

Bests

Jorge

[Jianjiong Gao]

Hi Jorge,

The values visualized in IGV are mean log values of segmented copy
number data. These values were used to calculated the discretized copy
number value (from -2 to 2) using GISTIC 2.0. The discretization is
calculated differently per sample -- it is not a simple round up or
taking a threshold. Please refer to GISTIC paper for details:
http://genomebiology.com/2011/12/4/R41

Best,
-JJ

> --
> You received this message because you are subscribed to the Google Groups
> "cBioPortal for Cancer Genomics Discussion Group" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to cbioportal+...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.