Figure 4From: Using logistic regression to improve the prognostic value of microarray gene expression data sets: application to early-stage squamous cell carcinoma of the lung and triple negative breast carcinoma Genes that predict early death or long survival with an accuracy of 0.8 or better. Sequential sets of 20 early death cases (1 – 20, 2 – 21,…21 – 40) were compared to a constant set of the 20 longest survivors (cases 88 – 107). The Y-axis denotes each sequential window; each row of dots indicates the genes which were found to be predictive for that window. Shown on the X-axis are the 45 genes most frequently identified of the 99 genes found to be predictive in one or more windows.Back to article page