10.1371/journal.pcbi.1007684.g005 Joël R. Pradines Joël R. Pradines Victor Farutin Victor Farutin Nicholas A. Cilfone Nicholas A. Cilfone Abouzar Ghavami Abouzar Ghavami Elma Kurtagic Elma Kurtagic Jamey Guess Jamey Guess Anthony M. Manning Anthony M. Manning Ishan Capila Ishan Capila Comparison of WAP and DEG scores reproducibility in eighteen large gene expression data sets representing eight diseases. Public Library of Science 2020 concept data sets precision medicine studies protein interaction networks gene expression data prioritize novel targets DEG gene expression analysis discovery rate estimation well-associated protein Identification WAP genome-wide transcriptional studies reproducibility drug target selection 2020-02-14 18:33:29 Figure https://plos.figshare.com/articles/figure/Comparison_of_WAP_and_DEG_scores_reproducibility_in_eighteen_large_gene_expression_data_sets_representing_eight_diseases_/11857398 <p>A-H: Each data set is randomly split in two sets one thousand times, and the resulting distributions of overlap statistic maxR are estimated for WAP scores (blue), PFP values of WAPs (pink), DEG scores (green) and WAP scores with randomly rewired protein network (brown). Vertical bars display 99% confidence intervals on median values estimated by bootstrapping with 10<sup>6</sup> samples. I: summary of results over 18 data sets for distributions of maxR differences between DEG and WAP scores (paired values of maxR in each partition (WAP versus DEG scores)).</p>