%0 Figure %A Pradines, Joël R. %A Farutin, Victor %A Cilfone, Nicholas A. %A Ghavami, Abouzar %A Kurtagic, Elma %A Guess, Jamey %A Manning, Anthony M. %A Capila, Ishan %D 2020 %T Reproducibility of WAPs and DEGs among four SSc data sets. %U https://plos.figshare.com/articles/figure/Reproducibility_of_WAPs_and_DEGs_among_four_SSc_data_sets_/11857392 %R 10.1371/journal.pcbi.1007684.g004 %2 https://plos.figshare.com/ndownloader/files/21731712 %K concept %K data sets %K precision medicine studies %K protein interaction networks %K gene expression data %K prioritize novel targets %K DEG %K gene expression analysis %K discovery rate estimation %K well-associated protein Identification %K WAP %K genome-wide transcriptional studies %K reproducibility %K drug target selection %X

Four gene expression data sets comparing skin samples between SSc patients and non-SSc subjects—GSE58095 (1), GS32413 (2), GSE9285 (3) and GSE45485 (4)—are used to obtain WAP and DEG scores. Blue and green colors represent overlaps between top scoring WAPs and DEGs respectively. A-B: Overlap profiles for two pairs of data sets. Solid profiles correspond to genes ordered by a two-sided t-test and dashed profiles—to ordering which further accounts for fold change (a gene rank is the average of its t-statistic and fold-change ranks). Brown profiles display results for ten randomly rewired protein networks. C-D: Overlap statistics maxR and AUC for all six data set pairs, when genes are ordered by a two-sided t-test. For randomly rewired networks, displayed values are averages.

%I PLOS Computational Biology