Models and data for predicting isoform abundance from UTR sequence features and GTEx PC

Each RDS file contains a ranger object, trained on transcripts after removing those associated with the held-out genes in one of the five cross-validation folds. The day and replicate number in the file name corresponds to the neuronal differentiation sample on which the model was trained. The file gene_folds.txt indicates the fold from which each gene was excluded during model training. The file transcript_gene_associations.txt contains transcript-gene associations. The file predictors.RDS contains the matrix of predictor variables. The file gtex_PCA_output.csv.gz contains the PC scores for each transcript in the GTEx data (needs to be extracted by gunzip).