Skip to main content

Table 1 Hazard ratio and logrank (p-value) for evaluated cut points in training, verification and validation datasets

From: A standardized microarray assay for the independent gene expression markers in AML: EVI1 and BAALC

Gene name Dataset AMLprofiler 25 thpercentile 50 thpercentile 75 thpercentile
BAALC Training .48 (.001) .44 (< .001) .63 (.01) .43(< .001)
Verification *) .60 (.002) .55 (.001) .66(.005) .65 (.007)
Validation .69 (.021) .72 (.066) .88 (.24) .93 (.345)
ERG Training No significant cutoff point 1.41 (.066) 1.37 (.051) 1.45 (.048)
Verification *) ND 1.41 (.037) 1.56 (.003) 1.72 (.001)
Validation ND 1.54 (.02) 1.12 (.248) 1.28 (.099)
MN1 Training 2.19 (< .001) 2.56 (< .001) 1.79 (.002) 1.79 (.003)
Verification *) 1.31 (.049) 1.12 (.26) 1.32 (.049) 1.39 (.035)
Validation 1.14 (.2329) 1.37 (.068) 1.19 (.167) 1.32 (.081)
  1. OS logrank p < 0.05 indicated in bold. For ERG no cut point was identified in the training cohort.
  2. *) this data set contains only normal karyotype cases.
  3. ND, not done because lack of significant cutoff point in training.