Xiaolei Zhang1,2, Roddy Walsh1,2, Nicola Whiffin1,2, Rachel Buchan1,2, William Midwinter1,2, Alicja Wilk1,2, Risha Govind1,2, Nicholas Li2,3, Mian Ahmad1,2, Francesco Mazzarotto1,4,5, Angharad Roberts1,2, Pantazis I Theotokis1,2, Erica Mazaika1,2, Mona Allouba1,6, Antonio de Marvao3, Chee Jian Pua7, Sharlene M Day8, Euan Ashley9, Steven D Colan10, Michelle Michels11, Alexandre C Pereira12, Daniel Jacoby13, Carolyn Y Ho14, Iacopo Olivotto4, Gunnar T Gunnarsson15, John L Jefferies16, Chris Semsarian17,18, Jodie Ingles17, Declan P O'Regan3, Yasmine Aguib1,6, Magdi H Yacoub1,6, Stuart A Cook1,2,7,19, Paul J R Barton1,2, Leonardo Bottolo20,21,22, James S Ware23,24,25. 1. National Heart and Lung Institute, Imperial College London, London, United Kingdom. 2. Cardiovascular Research Centre, Royal Brompton and Harefield NHS, Foundation Trust London, London, United Kingdom. 3. MRC London Institute of Medical Sciences, Imperial College London, London, United Kingdom. 4. Cardiomyopathy Unit, Careggi University Hospital, Florence, Italy. 5. Department of Clinical and Experimental Medicine, University of Florence, Florence, Italy. 6. Aswan Heart Centre, Magdi Yacoub Heart Foundation, Aswan, Egypt. 7. National Heart Centre, Singapore, Singapore. 8. Division of Cardiovascular Medicine and Penn Cardiovascular Institute, Perelman School of Medicine, University of Pennsylvania, Philadelphia, USA. 9. Division of Cardiovascular Medicine, Stanford University Medical Center, Stanford, CA, USA. 10. Department of Cardiology, Boston Children's Hospital, Boston, MA, USA. 11. Department of Cardiology, Thoraxcenter, Erasmus MC Rotterdam, Rotterdam, Netherlands. 12. Heart Institute (InCor), University of Sao Paulo Medical School, Sao Paulo, Brazil. 13. Department of Internal Medicine, Yale University, New Haven, CT, USA. 14. Cardiovascular Division, Brigham and Women's Hospital, Boston, MA, USA. 15. Faculty of Medicine, University of Iceland, Akureyri, Iceland. 16. The Cardiovascular Institute, University of Tennessee, Memphis, TN, USA. 17. Centenary Institute, The University of Sydney, Sydney, Australia. 18. Department of Cardiology, Royal Prince Alfred Hospital, Sydney, Australia. 19. Duke-National University of Singapore, Singapore, Singapore. 20. Department of Medical Genetics, University of Cambridge, Cambridge, United Kingdom. lb664@cam.ac.uk. 21. Alan Turing Institute, London, United Kingdom. lb664@cam.ac.uk. 22. MRC Biostatistics Unit, University of Cambridge, Cambridge, United Kingdom. lb664@cam.ac.uk. 23. National Heart and Lung Institute, Imperial College London, London, United Kingdom. j.ware@imperial.ac.uk. 24. Cardiovascular Research Centre, Royal Brompton and Harefield NHS, Foundation Trust London, London, United Kingdom. j.ware@imperial.ac.uk. 25. MRC London Institute of Medical Sciences, Imperial College London, London, United Kingdom. j.ware@imperial.ac.uk.
Abstract
PURPOSE: Accurate discrimination of benign and pathogenic rare variation remains a priority for clinical genome interpretation. State-of-the-art machine learning variant prioritization tools are imprecise and ignore important parameters defining gene-disease relationships, e.g., distinct consequences of gain-of-function versus loss-of-function variants. We hypothesized that incorporating disease-specific information would improve tool performance. METHODS: We developed a disease-specific variant classifier, CardioBoost, that estimates the probability of pathogenicity for rare missense variants in inherited cardiomyopathies and arrhythmias. We assessed CardioBoost's ability to discriminate known pathogenic from benign variants, prioritize disease-associated variants, and stratify patient outcomes. RESULTS: CardioBoost has high global discrimination accuracy (precision recall area under the curve [AUC] 0.91 for cardiomyopathies; 0.96 for arrhythmias), outperforming existing tools (4-24% improvement). CardioBoost obtains excellent accuracy (cardiomyopathies 90.2%; arrhythmias 91.9%) for variants classified with >90% confidence, and increases the proportion of variants classified with high confidence more than twofold compared with existing tools. Variants classified as disease-causing are associated with both disease status and clinical severity, including a 21% increased risk (95% confidence interval [CI] 11-29%) of severe adverse outcomes by age 60 in patients with hypertrophic cardiomyopathy. CONCLUSIONS: A disease-specific variant classifier outperforms state-of-the-art genome-wide tools for rare missense variants in inherited cardiac conditions ( https://www.cardiodb.org/cardioboost/ ), highlighting broad opportunities for improved pathogenicity prediction through disease specificity.
PURPOSE: Accurate discrimination of benign and pathogenic rare variation remains a priority for clinical genome interpretation. State-of-the-art machine learning variant prioritization tools are imprecise and ignore important parameters defining gene-disease relationships, e.g., distinct consequences of gain-of-function versus loss-of-function variants. We hypothesized that incorporating disease-specific information would improve tool performance. METHODS: We developed a disease-specific variant classifier, CardioBoost, that estimates the probability of pathogenicity for rare missense variants in inherited cardiomyopathies and arrhythmias. We assessed CardioBoost's ability to discriminate known pathogenic from benign variants, prioritize disease-associated variants, and stratify patient outcomes. RESULTS: CardioBoost has high global discrimination accuracy (precision recall area under the curve [AUC] 0.91 for cardiomyopathies; 0.96 for arrhythmias), outperforming existing tools (4-24% improvement). CardioBoost obtains excellent accuracy (cardiomyopathies 90.2%; arrhythmias 91.9%) for variants classified with >90% confidence, and increases the proportion of variants classified with high confidence more than twofold compared with existing tools. Variants classified as disease-causing are associated with both disease status and clinical severity, including a 21% increased risk (95% confidence interval [CI] 11-29%) of severe adverse outcomes by age 60 in patients with hypertrophic cardiomyopathy. CONCLUSIONS: A disease-specific variant classifier outperforms state-of-the-art genome-wide tools for rare missense variants in inherited cardiac conditions ( https://www.cardiodb.org/cardioboost/ ), highlighting broad opportunities for improved pathogenicity prediction through disease specificity.
Authors: Jodie Ingles; Charlotte Burns; Richard D Bagnall; Lien Lam; Laura Yeates; Tanya Sarina; Rajesh Puranik; Tom Briffa; John J Atherton; Tim Driscoll; Christopher Semsarian Journal: Circ Cardiovasc Genet Date: 2017-04
Authors: Chai-Ann Ng; Rizwan Ullah; Jessica Farr; Adam P Hill; Krystian A Kozek; Loren R Vanags; Devyn W Mitchell; Brett M Kroncke; Jamie I Vandenberg Journal: Am J Hum Genet Date: 2022-06-09 Impact factor: 11.043
Authors: Corey L Anderson; Saba Munawar; Louise Reilly; Timothy J Kamp; Craig T January; Brian P Delisle; Lee L Eckhardt Journal: Front Cardiovasc Med Date: 2022-07-04