Review 2: "Combining Blood Transcriptomic Signatures Improves the Prediction of Progression to Tuberculosis Among Household Contacts in Brazil"

The study presents valuable new data for the field of researchers developing prognostic and diagnostic transcriptomic biomarkers TB disease (data is available on request).TB cases were detected using Brazil national health records, looking for microbiologically confirmed active TB 4 years after household contacts of TB patients were consented and blood drawn for transcriptomic analysis; detection of cases in this way is rigorous, but may also be different than cases detected with active follow-up in a clinical trial.The authors start by evaluating previously published risk signatures in this new cohort.Two computational methods are used for this evaluation: GSVA and PLAGE.The two methods are appropriate for evaluating multi-gene signatures and PLAGE is particularly suitable for signatures with genes moving in opposite directions as it employs PCA for combing genes into a single score.GSVA is a nice complimentary method as it is non-