DrugReflector model checkpoints

DrugReflector is an ensemble of three multi-layer perceptron classifiers trained on Connectivity Map transcriptional signatures to predict compound classes from transcriptional signatures. The training set was partitioned into three non-overlapping splits, and each model was trained on two of the three splits using individual perturbation signatures as inputs (without replicate averaging). This DOI contains the resulting three torch model checkpoints, one for each split-trained model.