Name
Title | ||
|---|---|---|
A New Protein-Protein Interaction Prediction Algorithm Based on Conditional Random Field. |
Abstract | ||
|---|---|---|
It is very important to study on protein-protein interaction (PPI) because proteins carry out their cellular functions through their concerted interactions with other proteins. In order to overcome the shortcomings such as the strong independence assumptions and the label-bias problem in the existing methods for PPI prediction, we propose a novel method for PPI prediction in PPI network based on conditional random fields (CRF) model. The method transforms PPI prediction into the problem of sequential data labeling. Based on the properties of protein sequences, we design the methods of model constructing, training and decoding in CRF accordingly. Our method can determine the labeled sequence with maximum probability and thereby detect PPI. Experimental results on benchmark data sets show that our method is more accurate and efficient than other traditional methods. |
Year | Venue | Field |
|---|---|---|
2015 | ICIC | Protein–protein interaction prediction,Conditional random field,Sequential data,Data set,Random field,Pattern recognition,Computer science,Algorithm,Regular conditional probability,Artificial intelligence,Conditional entropy,Decoding methods |
DocType | Citations | PageRank |
Conference | 0 | 0.34 |
References | Authors | |
7 | 3 | |