Related Experiment Video
Updated: Jun 19, 2026
06:41
In Vivo Functional Study of Disease-associated Rare Human Variants Using Drosophila
Published on: August 20, 2019
Prediction of candidate primary immunodeficiency disease genes using a support vector machine learning approach
Shivakumar Keerthikumar1, Sahely Bhadra, Kumaran Kandasamy
1Institute of Bioinformatics, International Technology Park, Bangalore 560 066, India.
Summary
Identifying novel primary immunodeficiency disease (PID) genes is crucial. A new machine learning algorithm, trained on known PID genes, successfully predicted 1442 candidate genes, aiding in early disease identification.
Area of Science:
- Genetics
- Immunology
- Bioinformatics
Background:
- Early identification of primary immunodeficiency disease (PID) genes presents a significant clinical challenge.
- Existing resources catalog molecular alterations and phenotypes of known PID genes but lack predictive capabilities for novel candidates.
Purpose of the Study:
- To develop a predictive algorithm for identifying novel candidate primary immunodeficiency disease (PID) genes.
- To leverage the "Resource of Asian PDIs" platform for gene discovery.
Main Methods:
- Development of a machine learning algorithm using a support vector machine approach.
- Training the algorithm on 69 binary features from 148 known PID genes and 3162 non-PID genes.
- Utilizing the "Resource of Asian PDIs" platform data for feature extraction.
Main Results:
- Prediction of 1442 candidate PID genes.
- Experimental confirmation of six predicted genes as novel PID genes.
- Identification of a substantial list of potential PID genes for further investigation.
Conclusions:
- The developed machine learning algorithm effectively predicts candidate PID genes.
- This approach significantly aids in the discovery of new genes associated with primary immunodeficiency diseases.
- The predicted gene set offers valuable targets for etiological investigation in undiagnosed PID cases.
