Table 4 Classifier performance assessed via leave-one-batch-out cross validation
From: Blood-based identification of non-responders to anti-TNF therapy in rheumatoid arthritis
Left out data set | Treatment | Patient breakdown | AUROC(95 % CI) | Specificity(95 % CI) | Sensitivity(95 % CI) | Precision(95 % CI) | Likelihood ratio(95 % CI) |
|---|---|---|---|---|---|---|---|
GSE12051 | Infliximab | 7 NR, 37 R | 82 %**(66-98 %) | 97 %(86-100 %) | 14 %(0-58 %) | 50 %(1-99 %) | 5.29(0.4-74.9) |
GSE19821 | Infliximab | 5 NR, 10 R | 90 %**(74-100 %) | 100 %(59-100 %) | 40 %(5-85 %) | 100 %(9-100 %) | ∞(N/A) |
GSE58795 | Infliximab | 7 NR, 23 R | 57 %(26-87 %) | 91 %(72-99 %) | 43 %(10-82 %) | 60 %(15-95 %) | 4.93(1.0-23.8) |
GSE33377 | Infliximab | 13 NR, 14 R | 57 %(34-80 %) | 86 %(57-98 %) | 31 %(9-61 %) | 67 %(22-96 %) | 2.15(0.5-9.9) |