Ranking factors involved in diabetes remission after bariatric surgery using machine-learning integrating clinical and genomic biomarkers.

Adding information for the eight selected SNPs improved performance to an accuracy of 84% (AUC=0.92) and resulted in highly significant performance improvement as calculated by net reclassification improvement (NRI, NRI categorical: P value=1.45×10 −4 , NRI continuous: 2.81×10 −29 ) and integrated discrimination improvement ( P value=7.33×10 −25 ; Supplementary Table S1 ), emphasising the potential of including these genomic markers (see also receiver operating characteristic in Figure 3c ).

also in 132,142 other papers