I am exploring the PANR package but I don't understand the input data required. The vignette uses its own data which is described as "a matrix of NN z-scores with rows and columns corresponding to 249 TCs and seven reference TCs"

NN comes from Neural Network and TC is Treatment Conditions, however it is isn't explained how to calculate this from expression data, or why later the data used has 273 rows instead of the "corresponding" 249 NN z-scores.

Which data format should be used and how could I prepare it from expression data?