Back to Search
Start Over
Predicting pathogenicity of missense variants with weakly supervised regression
- Source :
- Hum Mutat
- Publication Year :
- 2019
- Publisher :
- Hindawi Limited, 2019.
-
Abstract
- Quickly growing genetic variation data of unknown clinical significance demand computational methods that can reliably predict clinical phenotypes and deeply unravel molecular mechanisms. On the platform enabled by CAGI (Critical Assessment of Genome Interpretation), we develop a novel "weakly supervised" regression (WSR) model that not only predicts precise clinical significance (probability of pathogenicity) from inexact training annotations (class of pathogenicity) but also infers underlying molecular mechanisms in a variant-specific fashion. Compared to multi-class logistic regression, a representative multi-class classifier, our kernelized WSR improves the performance for the ENIGMA Challenge set from 0.72 to 0.97 in binary AUC (Area Under the receiver operating characteristic Curve) and from 0.64 to 0.80 in ordinal multi-class AUC. WSR model interpretation and protein structural interpretation reach consensus in corroborating the most probable molecular mechanisms by which some pathogenic BRCA1 variants confer clinical significance, namely metal-binding disruption for C44F and C47Y, protein-binding disruption for M18T, and structure destabilization for S1715N. Availability: Source codes and data are provided at https://github.com/Shen-Lab/WSR-PredictPofPathogenicity/
- Subjects :
- Computer science
Mutation, Missense
Computational biology
Biology
Logistic regression
Genome
Article
Machine Learning
03 medical and health sciences
0302 clinical medicine
Genetics
Humans
Missense mutation
Genetic Predisposition to Disease
Clinical significance
Genetics (clinical)
030304 developmental biology
0303 health sciences
Models, Genetic
Receiver operating characteristic
BRCA1 Protein
030305 genetics & heredity
Computational Biology
Pathogenicity
Regression
Logistic Models
Phenotype
Area Under Curve
Classifier (UML)
030217 neurology & neurosurgery
Subjects
Details
- ISSN :
- 10981004 and 10597794
- Volume :
- 40
- Database :
- OpenAIRE
- Journal :
- Human Mutation
- Accession number :
- edsair.doi.dedup.....1cbb9909b57c2b2d581bd9452ada63a3