Table of Contents Author Guidelines Submit a Manuscript
The Scientific World Journal
Volume 2012 (2012), Article ID 380495, 11 pages
http://dx.doi.org/10.1100/2012/380495
Research Article

Gene Expression Profiles for Predicting Metastasis in Breast Cancer: A Cross-Study Comparison of Classification Methods

1Research Unit of Human Genetics, Institute of Clinical Research, University of Southern Denmark, Sdr. Boulevard 29, 5000 Odense C, Denmark
2Department of Clinical Genetics, Odense University Hospital, Sdr. Boulevard 29, 5000 Odense C, Denmark
3Institute of Public Health, University of Southern Denmark, J. B. Winsløws Vej 9B, 5000 Odense C, Denmark

Received 25 August 2012; Accepted 2 October 2012

Academic Editors: M. A. Kon and K. Najarian

Copyright © 2012 Mark Burton et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Supplementary Material

Supplementary Table 1 shows list of 283 rank-significant genes for classifier building, by gene symbol (left column) and by description (right column).

Supplementary Table 2 shows the internal validation results within the AM and RO datasets and the mean of AM & RO. These results are shown with respect to sensitivity (Sen), specificity (Spe) and balanced accuracy (bAcc) based on the ten times repeated ten folds cross-validation by each of the classification methods used. The classification are as follows: random forest (RF), logistic regression (LR), support vector machines with a radial (R-SVM), linear (L-SVM), polynomial (P-SVM), sigmoid kernel (S-SVM), a neural network with a single hidden layer (NNET) or cross method voting (Voting).

Supplementary Table 3 shows the testing results by classifiers developed in the RO dataset and validated in TR or MA and the combined mean performance in TR and MA. These results are shown with respect to sensitivity (Sen), specificity (Spe) and balanced accuracy (bAcc), by each of the classification methods used. The classification are as follows: random forest (RF), logistic regression (LR), support vector machines with a radial (R-SVM), linear (L-SVM), polynomial (P-SVM), sigmoid kernel (S-SVM), a neural network with a single hidden layer (NNET) or cross method voting (Voting).

Supplementary Table 4 shows the testing results by classifiers developed in the AM dataset and validated in TR or MA and the combined mean performance in TR and MA. These results are shown with respect to sensitivity (Sen), specificity (Spe) and balanced accuracy (bAcc), by each of the classification methods used. The classification are as follows: random forest (RF), logistic regression (LR), support vector machines with a radial (R-SVM), linear (L-SVM), polynomial (P-SVM), sigmoid kernel (S-SVM), a neural network with a single hidden layer (NNET) or cross method voting (Voting).

  1. Supplementary Material