false
zh-CN,zh-TW,en,fr,de,ja,ko,pt,es,th,vi
Catalog
2023 ACMG Annual Clinical Genetics Meeting Digital ...
Can Computational Tools Generate Greater Than ACMG ...
Can Computational Tools Generate Greater Than ACMG Supporting Evidence? A Circularity-Free Analysis Using ATM, CHEK2, and Breast Cancer Case-Control Data
Back to course
Pdf Summary
A study was conducted to analyze the effectiveness of computational tools in identifying the pathogenicity of rare missense substitutions. The 2015 ACMG guidelines recommend using computational tools to provide supporting evidence for or against pathogenicity. However, the study found that the predictive power of these tools may be distorted due to dependencies between tool scoring and prior variant classifications.<br /><br />The analysis involved case-control ATM and CHEK2 germline variant data from different populations. Four computational tools were used to score the missense substitutions, and the ATM gene was stratified into two regions. Frequentist odds ratios were calculated, and the results were correlated with ACMG classification bins.<br /><br />The study found a strong trend of increasing odds ratios across the ordered score ranges for all tested tools. All four tools provided Pathogenic_Moderate evidence, and two meta-predictors (BayesDel and REVEL) produced Pathogenic_Moderate_Plus evidence with Odds_Path above 9.0:1. The results showed a 4-step progression from Pathogenic_Moderate to Benign_Supporting based on the geometric averages of meta-predictor scores.<br /><br />Overall, the study suggests that computational tools can provide strong evidence in favor of pathogenicity for rare missense substitutions. The analysis used case-control data to calculate odds ratios, avoiding the dependencies introduced by the re-call approach. These findings contribute to the understanding of the true predictive power of computational tools in variant classification.
Asset Subtitle
Presenting Author - Julie L. Boyle, MS; Co-Author - Siqi Hu, MD; Co-Author - Scott Pew, MPH; Co-Author - Colin C. Young, PhD; Co-Author - Bing-Jian Feng, PhD; Co-Author - Colin B. Mackenzie, MD; Co-Author - Mia Hashibe, PhD; Co-Author - Sean V. Tavtigian, PhD;
Meta Tag
Bioinformatics
Co-Author
Siqi Hu, MD
Co-Author
Scott Pew, MPH
Co-Author
Colin C. Young, PhD
Co-Author
Bing-Jian Feng, PhD
Co-Author
Colin B. Mackenzie, MD
Co-Author
Mia Hashibe, PhD
Co-Author
Sean V. Tavtigian, PhD
Presenting Author
Julie L. Boyle, MS
Keywords
computational tools
pathogenicity
missense substitutions
ACMG guidelines
odds ratios
meta-predictors
variant classification
geometric averages
case-control data
predictive power
© 2024 American College of Medical Genetics and Genomics. All rights reserved.
×