Test for genes whose expression value is strongly predictive of batch (based on ROC classification). Important note: Assumes that the 'batch' of each cell is assigned to the cell's identity class (will be improved in a future release)
BatchGene(object, idents.use, genes.use = NULL, auc.cutoff = 0.6,
thresh.use = 0)
Seurat object
Batch names to test
Gene list to test
Minimum AUC needed to qualify as a 'batch gene'
Limit testing to genes which show, on average, at least X-fold difference (log-scale) in any one batch
Returns a list of genes that are strongly correlated with batch.