Sophie Belman1,2, Michael T Parsons3, Amanda B Spurdle3, David E Goldgar1, Bing-Jian Feng4. 1. University of Utah, Salt Lake City, UT, USA. 2. Wellcome Sanger Institute, University of Cambridge, Hinxton, UK. 3. QIMR Berghofer Medical Research Institute, Brisbane, QLD, Australia. 4. University of Utah, Salt Lake City, UT, USA. bingjian.feng@hsc.utah.edu.
Abstract
PURPOSE: The American College of Medical Genetics and Genomics (ACMG) and the Association for Molecular Pathology (AMP) have developed guidelines for classifying germline variants as pathogenic or benign to interpret genetic testing results. Cosegregation analysis is an important component of the guidelines. There are two main approaches for cosegregation analysis: meiosis counting and Bayes factor-based quantitative methods. Of these, the ACMG/AMP guidelines employ only meiosis counting. The accuracy of either approach has not been sufficiently addressed in previous works. METHODS: We analyzed hypothetical, simulated, and real-life data to evaluate the accuracy of each approach for cancer-associated genes. RESULTS: We demonstrate that meiosis counting can provide incorrect classifications when the underlying genetic basis of the disease departs from simple Mendelian situations. Some Bayes factor approaches are currently implemented with inappropriate penetrance. We propose an improved penetrance model and describe several critical considerations, including the accuracy of cosegregation for moderate-risk genes and the impact of pleiotropy, population, and birth year. We highlight a webserver, COOL (Co-segregation Online, http://BJFengLab.org/ ), that implements an accurate Bayes factor cosegregation analysis. CONCLUSION: An appropriate penetrance model improves the accuracy of Bayes factor cosegregation analysis for high-penetrant variants, and is a better choice than meiosis counting whenever feasible.
PURPOSE: The American College of Medical Genetics and Genomics (ACMG) and the Association for Molecular Pathology (AMP) have developed guidelines for classifying germline variants as pathogenic or benign to interpret genetic testing results. Cosegregation analysis is an important component of the guidelines. There are two main approaches for cosegregation analysis: meiosis counting and Bayes factor-based quantitative methods. Of these, the ACMG/AMP guidelines employ only meiosis counting. The accuracy of either approach has not been sufficiently addressed in previous works. METHODS: We analyzed hypothetical, simulated, and real-life data to evaluate the accuracy of each approach for cancer-associated genes. RESULTS: We demonstrate that meiosis counting can provide incorrect classifications when the underlying genetic basis of the disease departs from simple Mendelian situations. Some Bayes factor approaches are currently implemented with inappropriate penetrance. We propose an improved penetrance model and describe several critical considerations, including the accuracy of cosegregation for moderate-risk genes and the impact of pleiotropy, population, and birth year. We highlight a webserver, COOL (Co-segregation Online, http://BJFengLab.org/ ), that implements an accurate Bayes factor cosegregation analysis. CONCLUSION: An appropriate penetrance model improves the accuracy of Bayes factor cosegregation analysis for high-penetrant variants, and is a better choice than meiosis counting whenever feasible.
Authors: Sandrine M Caputo; Etienne Rouleau; Elizabeth Santana Dos Santos; Amanda B Spurdle; Dirce M Carraro; Adrien Briaux; Melissa Southey; Giovana Torrezan; Ambre Petitalot; Raphael Leman; Philippe Lafitte; Didier Meseure; Keltouma Driouch; Lucy Side; Carole Brewer; Sarah Beck; Athalie Melville; Alison Callaway; Françoise Revillion; Maria A A Koike Folgueira; Michael T Parsons; Heather Thorne; Anne Vincent-Salomon; Dominique Stoppa-Lyonnet; Ivan Bieche Journal: NPJ Breast Cancer Date: 2022-01-17