Learning from positive examples when the negative class is undetermined- microRNA gene identification

Malik Yousef¹^,3 , Segun Jung¹^,2^,4 , Louise C Showe¹ and Michael K Showe¹

¹Systems Biology Division, The Wistar Institute, Philadelphia, PA 19104, USA

²School of Biomedical Engineering, Science and Health Systems, Drexel University, Philadelphia, PA 19104, USA

³Computer Science, The College of Sakhnin, Sakhnin, Israel

⁴Sackler Institute of Graduate Biomedical Sciences, N.Y.U School of Medicine, New York, NY 10016, USA

author email corresponding author email

Algorithms for Molecular Biology 2008, 3:2doi:10.1186/1748-7188-3-2


Published:	28 January 2008

Additional files

Additional File 1:

Annotation of species used and additional data on accuracy associated with various one-class parameters. Table A. Sensitivity (Sen) and specificity (Spe) from one-class SVM using various word-lengths and the first 9 nt of the mature miRNA. Table B. Sensitivity (Sen) and specificity (Spe) obtained from one-class SVM to find the optimal number of the first k nucleotides using word length 3/4 3. Table C. Importance of the sequence features alone for classification. Table D. Optimized parameters for each one-class method. Table E. Annotation for all used species. Table F. The size of each dataset after removing similar structures of mature microRNAs. Table G. Accuracy in classification of All-miRNA dataset after masking to remove homologs. Table H. One-Class results obtained from the secondary features only and secondary features plus sequence features

Format: DOC Size: 178KB Download file

This file can be viewed with: Microsoft Word Viewer