170 -0.107 -0.232 18817 AL161983   -0.015 0.007 -0.037 17540 NM_016613 LOC51313

-0.002 0.022 -0.026 1723 AL133074   -0.078 -0.033 -0.123 23117 Contig14284_RC   -0.324 -0.209 -0.440 57 Contig56678_RC   -0.205 -0.135 -0.274 18904 NM_000125 ESR1 -0.312 -0.215 -0.409 6709 Contig57480_RC LOC51028 -0.021 0.009 -0.051 6105 NM_005113 GOLGA5 -0.046 -0.024 -0.067 To learn whether this gene signature could accurately predict survival of the patients from which it was created, we used our 20 gene signature to rank all 144 patients within the training set and divided them into a poor-prognosis group and good-prognosis group (Fig. 1A). We also compared the overall survival between the two groups (Fig. 1B, log-rank test[7], p < 0.0001), fitted linear regression to examine the correlation between time-to-death or censure and prognosis score (Fig.

1C, F-test, learn more significant negative correlation, p < 0.0001), and mean survival time SRT2104 cell line (or time to censure) between the two groups (Fig. 1D, Mann-Whitney test, p < 0.0001). In total, our results demonstrated the capacity of our gene signature to properly segregate human breast cancer patients into good- and poor-prognosis groups. Figure 1 Our 20-gene signature separates the training data set into poor-prognosis and good-prognosis groups (A, red = high expression, green = low expression) with differences in survival (B), a negative correlation between prognosis score and survival time (C) and differences in mean survival time (D). To validate our signature in patients whose

Metabolism inhibitor data had not been used to generate the signature, we divided the 151 patient validation group into poor-prognosis and good-prognosis groups (Fig. 2A). Again, our signature correctly separated patients based on survival (Fig. 2B, log-rank test p < 0.0001), correlated prognosis score with survival time (Fig. 2C, F-test, significant negative correlation, p = 0.034), and predicted Casein kinase 1 mean survival time (Fig. 2D, Mann-Whitney test, p = 0.0056). To rule out the possibility that our signature’s significance was a result of chance, we randomly generated a different 20-gene signature. As expected the random 20-gene signature did not separate patients into groups with differences in survival (Fig. 2E). Figure 2 Our 20-gene signature separates the validation data set into poor-prognosis and good-prognosis groups (A, red = high, green = low) with differences in survival (B), negative correlation between prognosis score and survival time (C), and differences mean survival time (D). E) A randomly generated 20-gene signature does not correlate prognosis score to patient survival. Analysis of the 20-gene signature To ensure that our algorithm produced predictors with comparable predictive power to other forms of feature selection we compared the 20-gene signature to a previously published Aurora kinase A expression model, as well as the FDA approved 70-gene signature (MammaPrint™) [2, 8].