A Regression-based K nearest neighbor algorithm for gene function prediction from heterogeneous data

Zizhen Yao¹ and Walter L Ruzzo¹^,2

¹Department of Computer Science and Engineering, AC101 Paul G. Allen Center, University of Washington, Seattle WA 98195, USA

²Department of Genome Sciences, University of Washington, Seattle WA 98195, USA

author email corresponding author email

BMC Bioinformatics 2006, 7(Suppl 1):S11doi:10.1186/1471-2105-7-S1-S11


Published:	20 March 2006

Additional files

Additional File 3:

Comparison of naive KNN methods and RBKNN methods on all combinations of data sources

Format: TXT Size: 2KB Download file

Additional File 1:

Comparison of ROC curves for COG. The axis at the right side represents the number of true positives, and the axis at the top represents the number of false positives. The plot also shows the black straight line for TP = FP. The intersection of this line with each ROC curve indicates the corresponding sensitivity and false positive rate at 50% accuracy. We only plot the region where FP rate is smaller than 0.15.

Format: PDF Size: 35KB Download file

This file can be viewed with: Adobe Acrobat Reader

Additional File 2:

Comparison of ROC curves for Multifunc. The axis at the right side represents the number of true positives, and the axis at the top represents the number of false positives. The plot also shows the black straight line for TP = FP. The intersection of this line with each ROC curve indicates the corresponding sensitivity and false positive rate at 50% accuracy. We only plot the region where FP rate is smaller than 0.20.

Format: PDF Size: 29KB Download file

This file can be viewed with: Adobe Acrobat Reader