Rare Variant Genotyping With SVM-Based Heterozygous Call Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Genotyping rare genetic variants is challenging due to their sparse representation in microarray platforms, leading to inaccurate heterozygous genotype calls, which complicates data analysis in clinical and research settings.
Innovation Solution
A method involving a microarray-based genotyping approach that includes a rare heterozygous adjustment (RHA) algorithm and a support vector machine (SVM) prediction model to improve the accuracy of rare heterozygous genotype calls by evaluating probe intensities and cluster distributions, setting false positives to 'No Call', and confirming true positives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If microarray platforms are used for genotyping rare variants, then high throughput and efficiency are achieved, but accuracy of heterozygous genotype calls deteriorates due to sparse representation
Solution Approach 1:
The method performs preliminary actions by first identifying heterozygous genotype calls using standard microarray clustering algorithms, then systematically evaluating each rare heterozygous call through multiple criteria (cluster position, probe intensity ratios, distance to cluster center) before final confirmation. This preliminary filtering and evaluation process resolves the contradiction by maintaining high throughput while improving accuracy through pre-screening mechanisms.
Solution Approach 2:
The invention implements feedback mechanisms by using support vector machine (SVM) prediction models that learn from training data to predict whether rare heterozygous calls are true positives or false positives. The system continuously refines its predictions based on multiple features including cluster characteristics and probe intensities, providing feedback-driven accuracy improvement without sacrificing throughput.
2Extent of automation
If clustering algorithms are used to identify genotypes for rare variants, then automated genotype calling is achieved, but reliability deteriorates because rare variants have no other samples in the heterozygous cluster
Solution Approach 1:
The invention introduces intermediary evaluation steps between automated clustering and final genotype confirmation. These intermediaries include assessing cluster position relative to homozygous clusters, evaluating probe intensity ratios, calculating distances to cluster centers, and applying SVM prediction models. These intermediary checks provide the reliability needed when direct clustering evidence is insufficient.
Solution Approach 2:
The method changes multiple parameters to assess rare heterozygous calls: cluster position coordinates, probe intensity ratios, distance metrics to cluster centers, and SVM prediction probabilities. By monitoring multiple parameters simultaneously, the system achieves reliable automated calling even when rare variants lack sufficient replicate samples in the heterozygous cluster.
3Ease of operation
If standard genotyping methods are used for rare variants, then simplicity of analysis is maintained, but false positive heterozygous calls increase
Solution Approach 1:
The invention segments the genotyping analysis into distinct phases: initial automated clustering, rare heterozygous candidate identification, multi-criteria evaluation (cluster position, probe intensities, distances), SVM prediction, and final confirmation. This segmentation maintains ease of operation by automating each segment while systematically reducing false positives through cumulative filtering at each stage.
Solution Approach 2:
The method performs preliminary filtering actions before final genotype confirmation, including assessing cluster characteristics, evaluating probe intensity ratios, calculating distance metrics, and applying SVM prediction models. These preliminary actions systematically eliminate false positives while maintaining analytical simplicity through automation.
Data Source
AI summary
A method of genotyping one or more genetic rare variants in a plurality of nucleic acid samples is described. The method comprises running one or more assays of the plurality of nucleic acid samples using at least one microarray comprising a plurality of probesets and generating assay data. Heterozygous genotypes in the plurality of nucleic acid samples are called and evaluated to identify any rare heterozygous genotype calls that are from a probeset having at least one probe, wherein the number of heterozygous genotype calls for the probeset does not exceed a maximum threshold value. Each identified rare heterozygous genotype call is evaluated to determine whether the identified rare heterozygous genotype call comprises a true rare heterozygous genotype call using a support vector machine prediction model or supervised machine learning classification model comprising a plurality of predictor values for identifying a true rare heterozygous genotype call.