How To Identify The Origin Of Tumor By Rnaseq
2
3
Entering edit mode
10.9 years ago

We have done RNAseq to unknown tumors in chicken. They tumors are are likely to be lymphoid tumors. I want to identify the tumor cell of origin (B-cell, T-cell, dendritic cell, macrophage,...). I tried 2 approaches: 1) I Identified 262 annotated genes with high FPKM ( > 200 as an arbitrary cut off) then I did pathway enrichment analysis using the available orthogonal genes in human (only 244 out of the initial 262). I did not have any pathognomonic signature, The lymphoid makers appears only in the low covered genes (FPKM <100). The group of gene of high coverage has a large no of ribosomal proteins. Any suggestions on this will be appreciated.

2) The second approach I tried was my trial to implement the "DeconRNASeq" R package (http://www.bioconductor.org/packages/2.12/bioc/html/DeconRNASeq.html) The package requires an expected expression profile of every cell type suspected to be seen in the heterogeneous tumor tissue. Is there a source of such expression profiles from different primary immune cells?

Thank you

cancer rnaseq • 4.3k views
ADD COMMENT
0
Entering edit mode

If you look in the pathology literature, is there a consensus on cell surface markers for your malignancy of interest? Even in blood tumors it's rare that expression is specific enough to determine the cell-of-origin when the possible candidates are closely related.

ADD REPLY
5
Entering edit mode
10.9 years ago

What I managed to do (for human cancer samples) is to get microarray data for various tissues (RNA-Seq-based illumine body map could also be used) and do a correlation analysis. Surprisingly, the correlations have shown clearly whether the tumor sample was from breast or prostate cancer. So I would recommend you get a list of well-annotated genes present on arrays for chicken and get FPKM for them. Then correlate log FPKM with log Intensity from arrays and see the closest tissue

ADD COMMENT
0
Entering edit mode

is there a user-friendly data base to get these microarray data from specially for tumors ?

ADD REPLY
0
Entering edit mode

The best thing I know is to browse the good old GEO ( http://www.ncbi.nlm.nih.gov/geo/ ).

ADD REPLY
2
Entering edit mode
10.9 years ago
IV ★ 1.3k

If you manage to get your hands on expression data such as mikhail suggested, I would have done PCA or hierarchical clustering based on gene expression.

Such techniques are easily implemented in R by most differential gene expression analysis packages, such as DESeq or EdgeR.

Cheers,

IV

ADD COMMENT

Login before adding your answer.

Traffic: 1748 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6