  Research articlePredicting prognosis using molecular profiling in estrogen receptor-positive breast cancer treated with tamoxifenSherene Loi1,2* 1 Functional Genomics Unit, Jules Bordet Institute, Brussels, Belgium 2 Peter MacCallum Cancer Center, East Melbourne, Victoria, Australia 3 Machine Learning Group, Université Libre de Bruxelles, Brussels, Belgium 4 NCCR Molecular Oncology, Swiss Institute of Cancer Research and Swiss Institute of Bioinformatics, Epalinges, Switzerland 5 Guys Hospital, London, UK 6 Molecular Cancer Genetics, Fondazione Istituto FIRC di Oncologia Molecolare (IFOM), Milan, Italy 7 Erasmus MC-Daniel-JNI, Rotterdam, The Netherlands 8 Department of Experimental Oncology, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy
BMC Genomics 2008, 9:239doi:10.1186/1471-2164-9-239
Additional filesAdditional file 1: Training dataset characteristics. Summary of patient and tumor characteristics of the 255 patients used in the development of the classifier. Format: XLS Size: 36KB Download file This file can be viewed with: Microsoft Excel Viewer Additional file 2: Probe clusters. (a) Full annotated gene list of all clusters (probe clusters; "pclust") used for classifier development after preliminary clustering and filtering on the separate dataset of 137 samples. (b) Full annotated gene list of probe clusters (pclust) used in final classifier and global model. Format: XLS Size: 320KB Download file This file can be viewed with: Microsoft Excel Viewer Additional file 3: Independent validation datasets characteristics. Summary of patient and tumor characteristics of the four independent validation sets. Format: XLS Size: 80KB Download file This file can be viewed with: Microsoft Excel Viewer Additional file 4: Signature size stability. Evolution of Stab criterion with respect to the signature size using multiple 10-fold cross-validation. The vertical dashed line represents the stability of the ranking using 13 pclusts. Format: PDF Size: KB Download file This file can be viewed with: Adobe Acrobat Reader Additional file 5: Signature size performance. Evolution of the log2 hazard ratio with respect to the signature size using multiple 10-fold cross-validation. The vertical dashed line represents the performance of the classifier using 13 pclusts. The horizontal red dashed line represents the limit of statistical significance. Format: PDF Size: 41KB Download file This file can be viewed with: Adobe Acrobat Reader Additional file 6: External validation of the classifier (Ma, Reid and Jansen datasets). (a) Kaplan Meier curves for Ma et al. The risk group was defined by the classifier using a 50:50 cutoff. The two survival curves were not significantly different according to the logrank test (p-value of 0.1). (b) Kaplan Meier curves for Reid et al. The risk group was defined by the classifier using a 50:50 cutoff. The two survival curves were significantly different according to the logrank test (p-value of 0.05). (c) Kaplan Meier curves for Jansen et al. The risk group was defined by the classifier using a 50:50 cutoff. The two survival curves were significantly different according to the logrank test (p-value of 0.25). Format: PDF Size: 116KB Download file This file can be viewed with: Adobe Acrobat Reader Additional file 7: Simplified classifiers. Top correlated "simplified" classifiers of 13 genes (one probe set per cluster) and histogram displaying correlation between "simple" classifiers and final model. Format: XLS Size: KB Download file This file can be viewed with: Microsoft Excel Viewer |
|
|
|
 |  | Log on / register Feedback | Support | My details |
on Google Scholar
author email
corresponding author email|
|
| © 1999-2009 BioMed Central Ltd unless otherwise stated. Part of Springer Science+Business Media. |