[Correspondence] Biopsy transcriptome expression profiling: proper validation is key – Authors' reply

Jeff Reeve and Philip F Halloran pose three questions regarding the derivation and validation of our predictive set of 13 genes:1 they suggest “overfitting” occurred; they question the accuracy of our methods with respect to external validation cohorts; and they claim our gene set is no better than “any random selection of 13 of these 10 000 probe sets”.

from #AlexandrosSfakianakis via Alexandros G.Sfakianakis on Inoreader http://ift.tt/2kYO1ox
via IFTTT