Feature Selection Protocol(y, gene_expression, max_num_genes = 15): let gene_expression be the expression of the set of all genes let Genes be the names of all genes let y be the target essentiality values p_vals_pearson = p-vals of pearson test with all genes’ expression and target p_vals_spearman = p-vals of spearman test with all genes’ expression and target p_vals_chisquared = p-vals of chi-squared test with genes’ expression and target selected_pearson = genes with false discovery rate below 0.05 selected_spearman = genes with false discovery rate below 0.05 selected_chi_square = genes with false discovery rate below 0.05 for each selected set of genes take the top max_num_genes selected = union(selected_pearson, selected_spearman, selected_chi_square) return selected |