Predict a CD8 immune phenotype using RNA-Seq data.
Arguments
- df
a data.frame with each row corresponding to a single observation. This data.frame should have all the columns available in the cd8ip_features_distribution. Feature values should be based on the
DESeq2::vst()transformation.
Value
a data.frame() with class probabilities (prob.inflamed,
prob.excluded, prob.desert) and the most-likely class
predicted.class. The n_out_of_range column indicates how many out of
range features were in the sample. In case more than 6 features
are out of range, the predictions will be NAs.