Research article
On reliable discovery of molecular signatures
-
* Corresponding author: Roland Nilsson rnilsson@broad.mit.edu
1 Computational Biology, Department of Physics, Linköping University, SE58183 Linköping, Sweden
2 Unit of Computational Medicine, King Gustav V Research Institute, Department of Medicine, Karolinska Institutet, SE17176 Stockholm, Sweden
BMC Bioinformatics 2009, 10:38 doi:10.1186/1471-2105-10-38
Published: 29 January 2009Additional files
Additional file 1:
Proofs. This document provides proofs of uniqueness and optimality of the optimal signature S*.
Format: PDF Size: 62KB Download file
This file can be viewed with: Adobe Acrobat Reader
Additional file 2:
KFD and WV methods, and convergence with increasing sample size. This figure shows the results corresponding to Figure 4 for the Kernel Fisher Discriminant (A-B) and Weighted Voting classification methods (C-D). Also shown is the convergence of the bootstrap method for the SVM classifier (E), where power approaches 1 as sample size increases.
Format: PDF Size: 222KB Download file
This file can be viewed with: Adobe Acrobat Reader
Additional file 3:
Gene signature for the Alon data set. Excel file detailing the gene signature discovered by the bootstrap method using the SVM classifier. The corresponding signature from Recursive Features elimination is also provided for reference.
Format: XLS Size: 21KB Download file
This file can be viewed with: Microsoft Excel Viewer
Additional file 4:
Gene signature for the Golub data set. Excel file detailing the gene signature discovered by the bootstrap method using the SVM classifier. The corresponding signature from Recursive Features elimination is also provided for reference.
Format: XLS Size: 74KB Download file
This file can be viewed with: Microsoft Excel Viewer
Additional file 5:
Gene signature for the Singh data set. Excel file detailing the gene signature discovered by the bootstrap method using the SVM classifier. The corresponding signature from Recursive Features elimination is also provided for reference.
Format: XLS Size: 37KB Download file
This file can be viewed with: Microsoft Excel Viewer