Check FDR thresholds for the enrichment t-stats

[lcolladotor]

Note that right now 2,691 genes pass the filterByExpr() in TGE and 8,443 pass that in WG. So I think that maybe the FDR 5% mark might be at a higher p-value in TGE than what we see in WG. But we haven’t checked this.

Make a table with 14 rows (7 x 2) and 4 columns. The 7 rows will be the 7 pathology categories, x 2 for TGE and WG. The columns will be: type (WG or TGE), pathology category (7 options) pvalue, and FDR.

Then the cells for the pvalue column will be the max p-value to FDR 5% for the enrichment t-stats. The FDR one will be the actual FDR.

I suspect that some rows will have NAs since we don't have FDRs close to 5% in many of the enrichment t-stat tests.

The inputs will be:

• WG: https://github.com/LieberInstitute/Visium_IF_AD/blob/master/code/05_deploy_app_wholegenome/Visium_IF_AD_wholegenome_model_results_FDR5perc.csv
• TGE: https://github.com/LieberInstitute/Visium_IF_AD/blob/master/code/06_deploy_app_targeted/Visium_IF_AD_targeted_model_results_FDR5perc.csv