.png){kind=icon}
.png){kind=icon}
Return
Using Weighted Extreme Learning Machine Combined With Scale-Invariant Feature Transform to Predict Protein-Protein Interactions From Protein Evolutionary Information
.png){kind=icon}
PRE
AI
X
Z
H
Z
林
DOI：10.1109/TCBB.2020.2965919.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Protein-Protein Interactions (PPIs) play an irreplaceable role in biological activities of organisms. Althoughmany high-throughput methods are used to identify PPIs fromdifferent kinds of organisms, they have some shortcomings, such as high cost and time-consuming. To solve the above problems, computationalmethods are developed to predict PPIs. Thus, in this paper, we present amethod to predict PPIs using protein sequences. First, protein sequences are transformed into PositionWeightMatrix (PWM), in which Scale-Invariant Feature Transform(SIFT) algorithmis used to extract features. Then PrincipalComponent Analysis (PCA) is applied to reduce the dimension of features. At last, Weighted Extreme LearningMachine (WELM) classifier is employed to predict PPIs and a series of evaluation results are obtained. In ourmethod, since SIFTandWELMare used to extract features and classify respectively, we called the proposedmethod SIFTWELM. When applying the proposedmethod on threewell-known PPIs datasets of Yeast, Human andHelicobacter:pylori, the average accuracies of our method using five-fold cross validation are obtained as high as 94.83, 97.60 and 83.64 percent, respectively. In order to evaluate the proposed approach properly, we compare itwith Support VectorMachine (SVM) classifier and other recent-developedmethods in different aspects. Moreover, the training time of our method is greatly shortened, which is obviously superior to the previousmethods, such as SVM, ACC, PCVMZMand so on.
Keywords:
Protein-protein interactions
scale-invariant feature transform
weighted extreme learning machine
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
I
IF:
3.4
Papers:
3.3K
Citations:
6.4K
Organization
Cited Papers