Refinement of the clinical variant interpretation framework by statistical evidence and machine learning

Med ◽  
2021 ◽  
Author(s):  
Atsushi Takata ◽  
Kohei Hamanaka ◽  
Naomichi Matsumoto
Keyword(s):  
Machine Learning ◽  
Statistical Evidence ◽  
Variant Interpretation ◽  
Clinical Variant

scholarly journals MACE2K: A Text-Mining Tool to Extract Literature-based Evidence for Variant Interpretation using Machine Learning

2020 ◽  
Author(s):  
Samir Gupta ◽  
Shruti Rao ◽  
Trisha Miglani ◽  
Yasaswini Iyer ◽  
Junxia Lin ◽  
...  
Keyword(s):  
Machine Learning ◽  
Deep Learning ◽  
Text Mining ◽  
Genomic Medicine ◽  
Relevant Information ◽  
Biomedical Literature ◽  
Variant Interpretation ◽  
Learning Models ◽  
Mining Tool ◽  
Text Mining Tool

AbstractInterpretation of a given variant’s pathogenicity is one of the most profound challenges to realizing the promise of genomic medicine. A large amount of information about associations between variants and diseases used by curators and researchers for interpreting variant pathogenicity is buried in biomedical literature. The development of text-mining tools that can extract relevant information from the literature will speed up and assist the variant interpretation curation process. In this work, we present a text-mining tool, MACE2k that extracts evidence sentences containing associations between variants and diseases from full-length PMC Open Access articles. We use different machine learning models (classical and deep learning) to identify evidence sentences with variant-disease associations. Evaluation shows promising results with the best F1-score of 82.9% and AUC-ROC of 73.9%. Classical ML models had a better recall (96.6% for Random Forest) compared to deep learning models. The deep learning model, Convolutional Neural Network had the best precision (75.6%), which is essential for any curation task.

scholarly journals Recommendations for application of the functional evidence PS3/BS3 criterion using the ACMG/AMP sequence variant interpretation framework

2019 ◽  
Vol 12 (1) ◽  
Author(s):  
Sarah E. Brnich ◽  
◽  
Ahmad N. Abou Tayoun ◽  
Fergus J. Couch ◽  
Garry R. Cutting ◽  
...  
Keyword(s):  
Protein Function ◽  
Working Group ◽  
Steering Committee ◽  
Moderate Level ◽  
Sequence Variant ◽  
Variant Interpretation ◽  
Functional Assays ◽  
Expert Opinions ◽  
Clinical Variant ◽  
Normal Gene

Abstract Background The American College of Medical Genetics and Genomics (ACMG)/Association for Molecular Pathology (AMP) clinical variant interpretation guidelines established criteria for different types of evidence. This includes the strong evidence codes PS3 and BS3 for “well-established” functional assays demonstrating a variant has abnormal or normal gene/protein function, respectively. However, they did not provide detailed guidance on how functional evidence should be evaluated, and differences in the application of the PS3/BS3 codes are a contributor to variant interpretation discordance between laboratories. This recommendation seeks to provide a more structured approach to the assessment of functional assays for variant interpretation and guidance on the use of various levels of strength based on assay validation. Methods The Clinical Genome Resource (ClinGen) Sequence Variant Interpretation (SVI) Working Group used curated functional evidence from ClinGen Variant Curation Expert Panel-developed rule specifications and expert opinions to refine the PS3/BS3 criteria over multiple in-person and virtual meetings. We estimated the odds of pathogenicity for assays using various numbers of variant controls to determine the minimum controls required to reach moderate level evidence. Feedback from the ClinGen Steering Committee and outside experts were incorporated into the recommendations at multiple stages of development. Results The SVI Working Group developed recommendations for evaluators regarding the assessment of the clinical validity of functional data and a four-step provisional framework to determine the appropriate strength of evidence that can be applied in clinical variant interpretation. These steps are as follows: (1) define the disease mechanism, (2) evaluate the applicability of general classes of assays used in the field, (3) evaluate the validity of specific instances of assays, and (4) apply evidence to individual variant interpretation. We found that a minimum of 11 total pathogenic and benign variant controls are required to reach moderate-level evidence in the absence of rigorous statistical analysis. Conclusions The recommendations and approach to functional evidence evaluation described here should help clarify the clinical variant interpretation process for functional assays. Further, we hope that these recommendations will help develop productive partnerships with basic scientists who have developed functional assays that are useful for interrogating the function of a variety of genes.

scholarly journals Prioritizing genes for systematic variant effect mapping

2020 ◽  
Author(s):  
Da Kuang ◽  
Rebecca Truty ◽  
Jochen Weile ◽  
Britt Johnson ◽  
Keith Nykamp ◽  
...  
Keyword(s):  
Protein Function ◽  
Supplementary Information ◽  
Functional Testing ◽  
Variant Interpretation ◽  
Missense Variants ◽  
Clinical Variant ◽  
Uncertain Significance ◽  
Missense Variation ◽  
Variant Effect ◽  
Extra Weight

Abstract Motivation When rare missense variants are clinically interpreted as to their pathogenicity, most are classified as variants of uncertain significance (VUS). Although functional assays can provide strong evidence for variant classification, such results are generally unavailable. Multiplexed assays of variant effect can generate experimental ‘variant effect maps’ that score nearly all possible missense variants in selected protein targets for their impact on protein function. However, these efforts have not always prioritized proteins for which variant effect maps would have the greatest impact on clinical variant interpretation. Results Here, we mined databases of clinically interpreted variants and applied three strategies, each building on the previous, to prioritize genes for systematic functional testing of missense variation. The strategies ranked genes (i) by the number of unique missense VUS that had been reported to ClinVar; (ii) by movability- and reappearance-weighted impact scores, to give extra weight to reappearing, movable VUS and (iii) by difficulty-adjusted impact scores, to account for the more resource-intensive nature of generating variant effect maps for longer genes. Our results could be used to guide systematic functional testing of missense variation toward greater impact on clinical variant interpretation. Availability and implementation Source code available at: https://github.com/rothlab/mave-gene-prioritization Supplementary information Supplementary data are available at Bioinformatics online.

scholarly journals A calibrated functional patch clamp assay to enhance clinical variant interpretation in KCNH2-related long QT syndrome

2021 ◽  
Author(s):  
Connie Jiang ◽  
Ebony Richardson ◽  
Jessica Farr ◽  
Adam P Hill ◽  
Rizwan Ullah ◽  
...  
Keyword(s):  
Patch Clamp ◽  
Current Density ◽  
Hek293 Cells ◽  
Functional Assay ◽  
Sequence Variant ◽  
Supporting Evidence ◽  
Variant Classification ◽  
Variant Interpretation ◽  
Clinical Variant ◽  
Automated Patch Clamp

Purpose: Modern sequencing technologies have revolutionised our detection of gene variants. In most genes, including KCNH2, the majority of missense variants are currently classified as variants of uncertain significance (VUS). The aim of this study is to investigate the utility of an automated patch-clamp assay for aiding clinical variant classification in the KCNH2 gene. Methods: The assay was designed according to recommendations of the ClinGen sequence variant interpretation framework. Thirty-one control variants of known clinical significance (17 pathogenic/likely pathogenic, 14 benign/likely benign) were heterozygously expressed in Flp-In HEK293 cells. Variants were analysed for effects on current density and channel gating. A panel of 44 VUS was then assessed for reclassification. Results: All 17 pathogenic variant controls had reduced current density and 13/14 benign variant controls had normal current density, which enabled determination of normal and abnormal ranges for applying moderate or supporting evidence strength for variant classification. Inclusion of KCNH2 functional assay evidence enabled us to reclassify 6 out of 44 VUS as likely pathogenic. Conclusion: The high-throughput patch clamp assay can provide moderate strength evidence for clinical interpretation of clinical KCNH2 variants and demonstrates the value proposition for developing automated patch clamp assays for other ion channel genes.

scholarly journals Using large sequencing data sets to refine intragenic disease regions and prioritize clinical variant interpretation

2016 ◽  
Vol 19 (5) ◽  
pp. 496-504 ◽  
Author(s):  
Sami S. Amr ◽  
Saeed H. Al Turki ◽  
Matthew Lebo ◽  
Mahdi Sarmady ◽  
Heidi L. Rehm ◽  
...  
Keyword(s):  
Data Sets ◽  
Variant Interpretation ◽  
Sequencing Data ◽  
Clinical Variant

scholarly journals High-throughput PRPF31 variant characterisation pipeline consistent with ACMG/AMP clinical variant interpretation guidelines

2020 ◽  
Author(s):  
Liliya Nazlamova ◽  
Man-Kim Cheung ◽  
Jelmer Legebeke ◽  
Jenny Lord ◽  
Reuben J. Pengelly ◽  
...  
Keyword(s):  
Gene Therapy ◽  
High Throughput ◽  
Quantitative Measure ◽  
Dominant Negative ◽  
Retinal Cell ◽  
Variant Interpretation ◽  
Negative Effects ◽  
Missense Variants ◽  
Clinical Variant

AbstractMutations in PRPF31 are the second most common cause of the degenerative retinal condition autosomal dominant retinitis pigmentosa. Difficulty in characterising missense variants in this gene presents a significant challenge in providing accurate diagnosis for patients to enable targeted testing of other family members, aid family planning, allow pre-implantation diagnosis and inform eligibility for gene therapy trials. With PRPF31 gene therapy in development, there is an urgent need for tools for accurate molecular diagnosis. Here we present a high-throughput high content imaging assay providing quantitative measure of effect of missense variants in PRPF31 which meets the recently published criteria for a baseline standard in vitro test for clinical variant interpretation. This assay utilizes a new and well-characterized PRPF31+/- human retinal cell line generated using CRISPR gene editing, which allows testing of PRPF31 variants which may be causing disease through either haploinsufficiency or dominant negative effects, or a combination of both. The mutant cells have significantly fewer cilia than wild-type cells, allowing rescue of ciliogenesis with benign or mild variants, but do not totally lack cilia, so dominant negative effects can be observed. The results of the assay provide BS3_supporting evidence to the benign classification of two novel uncharacterized PRPF31 variants and suggest that one novel uncharacterized PRPF31 variant may be pathogenic. We hope that this will be a useful tool for clinical characterisation of PRPF31 variants of unknown significance, and can be extended to variant classification in other ciliopathies.

Sign in / Sign up

Export Citation Format

Share Document