Functional prediction

Prediction of protein function using protein-protein interaction data
Minghua Deng , Kui Zhang , Shipra Mehta , 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.

Abstract

Assigning functions to novel proteins is one of the most important problems in the post-genomic era. Several approaches have been applied to this problem, including analyzing gene expression patterns, phylogenetic profiles, protein fusions and protein-protein interactions. We develop a novel approach that applies the theory of Markov random fields to infer a protein's functions using protein-protein interaction data and the functional annotations of its interaction protein partners. For each function of interest and a protein, we predict the probability that the protein has that function using Bayesian approaches. Unlike in other available approaches for protein annotation where a protein has or does not have a function of interest, we give a probability for having the function. We apply our method to predict three functional categories for yeast proteins followed Yeast Proteome database functional category (YPD, http://www.incyte.com/), using the protein-protein interaction data from the Munich Information Center for Protein Sequences (MIPS, http://mips.gfs.de/). We show that our approach outperforms other available methods for function prediction based on protein interaction data.

Original data

Predictions

Biochemical function predictions
- Leave-one-out for known proteins Tex file (1.6M byte), Compressed file (134K byte--Recommended).
- For unknown proteins Tex file (540K byte), Compressed file (79K byte--Recommended).
Cellular role predictions
- Leave-one-out for known proteins Tex file (1.5M byte), Compressed file (137K byte--Recommended ).
- For unknown proteins Tex file (273K byte), Compressed file (44K byte--Recommended).
Subcellular location predictions
- Leave-one-out for known proteins Tex file (862K byte), Compressed file (74K Byte--Recommended).
- For unknown proteins Tex file (307K byte), Compressed file (49K byte--Recommended).

Tables and figures.

Data information, Table1.
Parameters for different functional categories.
- Biochemical function category, Table2.
- Cellular role category, Table3.
- Subcellular location category, Table4.
Sensitivity and specificity vs posterior probability cutoff based on different functional categories.
- Biochemical function category, Figure1.
- Cellular role category, Figure2.
- Subcellular location category, Figure3.
Figures: Sensitivity vs specificity based on different functional categories.
- Biochemical function category, Figure4.
- Cellular role category, Figure5.
- Subcellular location category, Figure6.
Figures: Sensitivity vs specificity for proteins with different partners based on different functional categories.
- Biochemical function category, Figure7.
- Cellular role category, Figure8.
- Subcellular location category, Figure9.
A interaction subnetwork and posterior probability for proteins based on 43th cellular role, "Vescular transport", Figure10.
Functions for proteins involved in above subnetwork, Table5

Fengzhu Sun

Last modified: Tue., July 23, 2002.