Integrated functional prediction

An Integrated Probabilistic Model for Functional Prediction of Proteins
Minghua Deng , Ting Chen and Fengzhu Sun
Molecular and Computational Biology Program, Department of Biological Sciences, University of Southern California, 1042 West 36th Place, Los Angeles, CA 90089-1113, USA.

Unchanged data.

Result based on expression data (10-fold cross-validation).

Spellman's cell cycle expression data.
Pearson correlation coefficients.
Highly correlated (Pearson correlation>=0.80) protein pairs: 1366.
Known proteins covered by at least one data source: 3588.
Unknown proteins covered by at least one data source: 1516.
Coverage detail for all proteins: 5104
10-fold cross-validation sample sets of partition A.
10-fold cross-validation results for samples in partition A.
10-fold cross-validation sample sets of partition B.
10-fold cross-validation results for samples in partition B.
10-fold cross-validation sample sets of partition C.
10-fold cross-validation results for samples in partition C.

Results based on expression data (5-fold cross-validation).

Now the coverage detail is the same as that in 10-fold cross-valiation. We simply merged two 10-fold sample sets into one 5-fold sample set. To garantee new sample sets having almost same size, we merged 1+6, 2+7, 3+8, 4+9, 5+10 for each partition
5-fold cross-validation sample sets of partition A.
5-fold cross-validation results for samples in partition A.
5-fold cross-validation sample sets of partition B.
5-fold cross-validation results for samples in partition B.
5-fold cross-validation sample sets of partition C.
5-fold cross-validation results for samples in partition C.

Result without expression data.

Known proteins covered by at least one data source: 3573.
Unknown proteins covered by at least one data source: 1473.
Coverage detail for all proteins: 5046
10-fold cross-validation sample sets of partition A.
10-fold cross-validation results for samples in partition A.
10-fold cross-validation sample sets of partition B.
10-fold cross-validation results for samples in partition B.
10-fold cross-validation sample sets of partition C.
10-fold cross-validation results for samples in partition C.

Minghua Deng

Created: Friday, May 09, 2003. Last modified: Friday, Jun 20, 2003.