Detecting Co-Dependencies via RNAi data

[Wenyi]

Hi DepMap team,

I just downloaded “RNAi_(Achilles+DRIVE+Marcotte,_DEMETER2).csv” from the Depmap portal via the “custom download” function. However, I found that it included many NA values. I would like to know what should I do with these NA in order to get similar results to the online analysis results in the Depmap portal. Should I manually remove all cell lines with NA or all genes with NA? After that, I wonder if the Pearson coefficient is suitable to get Co-Dependencies. Or, should I use Spearman coefficients to detect co-dependencies since gene expression does not seem to fit a normal distribution?
I would be very grateful for any advice you could give me.

Best,
Wenyi Jin

[pmontgom]

The code we use for generating co-dependencies is linked to in this post:

Down load of co-dependencies

Sure, I posted some code at https://gist.github.com/pgm/ac2ac4c664ef81200ce49133cc4cee02 This code is modified from the portal’s codebase which will compute the top N correlates for given gene effect matrix downloaded from the portal. Running the following would compute the top 10 co-dependencies: python scripts/correlation_from_csv.py Achilles_gene_effect.csv Achilles_gene_effect.csv --limit 10 out.csv As run above it will correlate the same matrix against itself, but one could also use the…

We use pairwise complete observations when computing correlation, so if one sample has a NA for a given gene, we do not include that sample when comparing each gene to that one.

Also, we use pearson correlation instead of spearman correlation. You can read more about the reasoning for why in this blog post: When not to use Spearman correlations - Cancer Data Science Blog

Thanks,
Phil