Some questions

[keremw]

Hi,
Thank you for this wonderful tool. I have a few questions regarding the input and output.

1. I understand that --match finds a drug that matches the DEG files used, but what does the option without the match do?
2. In the paper you've mentioned that the drugref includes both upregulation and downregulation ("The proto-matrix itself contains information including genes acted on by a specific drug and the directionality in which it is influenced, that is, whether the drug induces up or down regulation of the gene.") But in oppose to the Broad GSEA the NES includes only positive values and non negative values that indicate enrichment in downregulation. Is there a way to know if a DEG profile I am checking is upregulated as drug X. Or downregulated as drug X?
3. On the pms files for each gene what does a "0" or "1" stand for? Are genes marked as "upregulated" and 0 "downregulated? How does that interact with the enrichment score?
   Again,
   Thanks for your great tool,
   Kerem

[sxf296]

Hi Kerem,

Apologies for the extremely late response. I have since graduated from my program, and I rarely check this repo. To answer your questions:

1. Without the --match flag, the method will look for complementary expression regulation. This is what you want when you're looking for drugs to counteract DEGs.
2. Unfortunately, this is not explicit wrt to the output of the method. You can safely assume that all genes are being oppositely regulated by any particular drug in the results generated (by nature of the algorithm). This is especially true for the leading edge genes as listed in the results as those are the ones driving the ES. You can check your toptable for any particular driver gene and flip the sign for the drug effect on that gene for now.
3. From what I recall, I coded 0 as drug down-regs gene X and 1 as drug up-regs gene X. The enrichment score is based off of either a matching (--match) or opposite effect of this wrt to your DEGs.

In the future please email me any questions. It is a more effect way to reach me.

Thanks,
Mike

[echoduan]

Hi Fang,

 what's difference between CM_P20.csv and L1K_P20.csv in the pms folder ? Which one I had better use ? Look forward to your reply. Thanks

[sxf296]

For CM_P20.csv, CM implies that the data you are using to perform the analysis is generated from CMAP with signatures sizes of 20 genes ranked by most significant (p). L1K_P20.csv is the same but generated LINCS1000 data. Feel free to use either one. We recommend LINCS for a more extensive analysis.