Comparing NCI-60 transcriptome with another variable to determine genes that correlate

Question

[Adrian Pelin]

Hello, I will probably not formulate this properly but I want to compare the transcriptome of NCI-60 cell lines with the ability of my virus to kill these cell lines. Ideally, I want to find genes whose expression correlates with ability of virus to kill that cell.

There used to be a tool called COMPARE (https://dtp.cancer.gov/databases_tools/compare.htm), however the public version seems to glitch.

Is there a name for this statistical analysis?

Answer 1

[cannin]

You can use the rcellminer R package: http://bioconductor.org/packages/release/bioc/vignettes/rcellminer/inst/doc/rcellminerUsage.html#correlating-dna-copy-number-alteration-and-gene-expression

to quickly identify genes of interest that match a pattern of interest in the NCI-60.

Separately, the CellMinerCDB website: https://discover.nci.nih.gov/cellminercdb has a 'Compare Patterns' feature that lets users genes/drugs that correlate well with the expression of an input query gene.

Answer 2

[colindaven]

I would advise you to do this sort of analysis in R, if you can. Rstudio eases you into things a lot and would help with this sort of correlation.

If that's not an option you can look at public Galaxy instances and particularly the DeepTools package to look at multiple correlations across large datasets.

Principal components analysis is likely to help with your data exploration as well.