Table 2.

Normal and tumor breast tissue class predictor miRNAs

miRNA nameMedian expressionANOVA * PSVM prediction strength PAM score Chromosome map
CancerNormalCancerNormal
mir-009-11.361.010.00918.050.011−0.1021q22
mir-010b1.111.700.04498.70−0.0320.2992q31
mir-0211.671.080.004710.200.025−0.23517q23.2
mir-0341.671.090.01068.050.011−0.1061p36.22
mir-102 (mir-29b)1.361.14>0.108.920.000−0.0041q32.2-32.3
mir-123 (mir-126)0.921.130.09409.13−0.0150.1389q34
mir-125a1.201.730.00338.99−0.0400.38119q13.4
mir-125b-11.302.870.026514.78−0.0960.91511q24.1
mir-125b-21.262.630.023317.62−0.1061.00621q11.2
mir-140-as0.931.100.069511.01−0.0050.05016q22.1
mir-1451.523.610.004012.93−0.1581.5025q32-33
mir-155 (BIC)1.751.370.001210.920.003−0.03021q21
mir-1940.961.09>0.1011.12−0.0250.2341q41
mir-2040.780.890.00228.10−0.0150.1449q21.1
mir-2133.722.470.01089.440.023−0.2201q31.3-q32.1
  • * ANOVA (Welch t test in the Genespring software package) as calculated in Table 1.

  • Support Vector Machine prediction analysis tool (from Genespring 7.2 software package). Prediction strengths are calculated as negative natural log of the probability to predict the observed number of samples, in one of the two classes, by chance. The higher is the score, the best is the prediction strength.

  • Centroid scores for the two classes of the PAM ( 18).