lcolladotor
commented
3 years ago So it's like a table with 3 or 4 columns:
- Gene ID (or gene symbol if gene ID is not available)
- Group. Can be "direction": higher in controls or higher in cases in case of differentially expressed genes. But it could be another grouping variable.
- Study name.
- Optional: p-value. I mean optional since at https://github.com/LieberInstitute/HumanPilot/blob/master/Analysis/Layer_Guesses/check_clinical_gene_sets.R we filtered other lists to just the significant genes (like FDR < 0.05).
Create lists of ENSEMBL/GENCODE gene IDs (not gene Symbols since they are not unique) of genes related to AD (separating those that have higher expression in controls vs those that higher expression in cases) from the literature. Then we can use code like https://github.com/LieberInstitute/HumanPilot/blob/master/Analysis/Layer_Guesses/check_clinical_gene_sets.R to compute chi square tests like at http://research.libd.org/spatialLIBD/reference/gene_set_enrichment.html that we then use to make plots like http://research.libd.org/spatialLIBD/reference/gene_set_enrichment_plot.html