Tumor microenvironment in TCGA data

[Gorostiola Gonzalez, M.]

Dear sir/madam,

 

My name is Marina Gorostiola González. I am a PhD candidate in the group of Gerard van Westen (Computational Drug Discovery, LACDR, Leiden University, the Netherlands).

 

As part of my project, I am working with some TCGA data from the BROAD firehose.

 

Lately, my attention was drawn to the fact that tumor microenvironment information can be derived from tumor samples (e.g. RNAseq data) in the TCGA.

The idea would be that in hot tumors, where the immune system recruitment is high, the tumor samples obtained from patients and sequenced would also include immune cells from the tumor microenvironment. Therefore, by scanning for overexpression of immune cell-specific genes (e.g. CD4, CD8, CD25), one could determine from the RNAseq data in TCGA whether the tumor is hot or cold.

This arises two questions:

1)      Is this the case for TCGA samples? Or on the contrary tumor samples do not include non-tumoral cells?

2)      If this is indeed the case, should we be aware in such cases that not only RNAseq data, but also other sequencing data (e.g. genetic data), is not only representing the tumor cells themselves but also possibly immune cells? (I.e. in a “hot” tumor, where CD4 expression is high, if we see that a specific mutation is present in e.g. Adenosine receptor A1, we cannot say for sure whether this mutation happened in the tumor cells or in immune cells?).

 

Your answers to these questions would be very helpful for my research.

Thank you in advance and best regards,

 

Marina Gorostiola González
PhD Candidate

PI: Dr. Laura H. Heitman & Dr. Gerard J.P. van Westen

Division Drug Discovery and Safety
Leiden Academic Center for Drug Research, Leiden University
Einsteinweg 55, 2333 CC Leiden, The Netherlands
E-mail: m.gorostio...@lacdr.leidenuniv.nl