2023 ACMG Annual Clinical Genetics Meeting Digital Edition: Poster Gallery-Predicting variant classification by modeling the effects of functional and case study data in <em>RUNX1 </em>variants of uncertain significance

Predicting variant classification by modeling the effects of functional and case study data in RUNX1 variants of uncertain significance

Back to course

Pdf Summary

This study focuses on predicting the classification of variants in the RUNX1 gene based on functional and case study data. Currently, over 50% of deposited RUNX1 variants in ClinVar are considered variants of uncertain significance (VUS) due to a lack of information for classification. The primary means of further classifying these VUS is through functional evidence or identification of additional probands or family segregation, but this can be challenging due to the rarity of the disorder. <br /><br />The researchers aimed to identify potential RUNX1 variants that would be significantly impacted by the addition of hypothetical evidence to prioritize them for directed functional testing. They used a deep generative model called EVE to predict variant pathogenicity and determine which ACMG/AMP codes to apply for each variant. They defined a threshold EVE score to distinguish potentially benign versus pathogenic variants.<br /><br />Using this approach, all VUS variants can be reclassified with the addition of hypothetical functional and case study data. The researchers found that the Runt-homology Domain of the RUNX1 gene produces the most pathogenic variants compared to the rest of the gene. Functional testing was identified as the most targeted way to reclassify a variant.<br /><br />The researchers plan to use their methodology to prioritize RUNX1 VUS for inter-lab collaborative functional testing. They also aim to apply this in silico methodology to predict the classification of variants that have conflicting benign and pathogenic ACMG/AMP criteria. Additionally, they hope to apply this methodology to other genes.<br /><br />In conclusion, this study demonstrates the potential of using functional and case study data to reclassify VUS variants in the RUNX1 gene. This approach can help prioritize variants for further testing and improve variant classification accuracy.

Asset Subtitle

Presenting Author - Mancy Shah, MSc; Co-Author - Winslow Johnson, MS; Co-Author - Lucy A. Godley, MD, PhD; Co-Author - David Wu, MD, PhD;

Meta Tag

Bioinformatics

Counseling

Exome sequencing

Genome sequencing

Methodology

Risk Assessment

Sequencing

Co-Author Winslow Johnson, MS

Co-Author Lucy A. Godley, MD, PhD

Co-Author David Wu, MD, PhD

Presenting Author Mancy Shah, MSc

Keywords

RUNX1 gene

variant classification

variants of uncertain significance

functional evidence

case study data

deep generative model

EVE

pathogenicity prediction

ACMG/AMP codes

reclassification