DECIPHER: Improving Genetic Diagnosis Through Dynamic Integration of Genomic and Clinical Data

Abstract

DECIPHER (Database of Genomic Variation and Phenotype in Humans Using Ensembl Resources) shares candidate diagnostic variants and phenotypic data from patients with genetic disorders to facilitate research and improve the diagnosis, management, and therapy of rare diseases. The platform sits at the boundary between genomic research and the clinical community. DECIPHER aims to ensure that the most up-to-date data are made rapidly available within its interpretation interfaces to improve clinical care. Newly integrated cardiac case-control data that provide evidence of gene-disease associations and inform variant interpretation exemplify this mission. New research resources are presented in a format optimized for use by a broad range of professionals supporting the delivery of genomic medicine. The interfaces within DECIPHER integrate and contextualize variant and phenotypic data, helping to determine a robust clinico-molecular diagnosis for rare-disease patients, which combines both variant classification and clinical fit. DECIPHER supports discovery research, connecting individuals within the rare-disease community to pursue hypothesis-driven research.

Keywords: genetic diagnosis; genetic disorders; genotype–phenotype correlation; rare diseases; variant interpretation.

Figure 1. DECIPHER is a web-based platform that shares genotype and phenotype data from rare-disease patients.

(a) DECIPHER openly shares more than 44,000 patient records. (b) DECIPHER enables contact between depositing centers to facilitate research and improve diagnosis. Each line on the map represents a message sent between depositing centers.

Figure 2. DECIPHER provides easy access to clinical recommendations for assessing variant pathogenicity.

(a) The sequence variant interpretation interface allows users to assess pathogenicity according to ACMG/AMP guidelines. Links to ClinGen general recommendations and gene- or disease-specific recommendations are provided, in addition to ClinGen expert panel interpretations where available. (b) ClinGen gene- or disease-specific recommendations can be displayed in DECIPHER. The Rett and Angelman-Like Disorders VCEP example shown here has recommendations for the use of the PM1 criterion for CDKL5 disorder. (c) The CNV interpretation interface allows users to assess pathogenicity according to ACMG/ClinGen technical standards. Abbreviations: ACMG, American College of Medical Genetics and Genomics; AMP, Association for Molecular Pathology; ClinGen, Clinical Genome Resource; CNV, copy number variant; VCEP, variant curation expert panel.

Figure 3. DECIPHER provides CNV annotation to assist in variant interpretation.

(a) DECIPHER displays gene dosage sensitivity metrics in tables that list genes overlapping a patient’s CNV or structural variant to assist in the identification of candidate genes. Gene–disease associations from the OMIM Morbid Map, G2P, and ClinGen are also provided. (b) Dosage sensitivity scores (which are a predictor of pathogenicity) and sampling probabilities (which estimate the proportion of the general population that carry a rare deletion/duplication with a dosage sensitivity score that is as severe as or more severe than the score for the CNV being assessed) are displayed for each deposited CNV. (c) A graph is available for each CNV that displays the dosage sensitivity score of the CNV along with the control population. Abbreviations: ClinGen, Clinical Genome Resource; CNV, copy number variant; G2P, Gene2Phenotype; OMIM, Online Mendelian Inheritance in Man.

Figure 4

DECIPHER displays an Ensembl Regulatory Build track in the genome browser to assist interpretation of variants in the noncoding genome. The tracks displayed here are genes (colored by pHaplo score), regulatory features, and transcripts. The regulatory features available are promoters (regions at the 5’ end of genes where transcription factors and RNA polymerase bind to initiate transcription), promoter flanks (transcription factor binding regions that flank promoters), enhancers (regions that bind transcription factors and interact with promoters to stimulate transcription of distant genes), CTCF binding sites (regions that bind CTCF, the insulator protein that demarcates open and closed chromatin), transcription factor binding sites (sites that bind transcription factors, for which no other role can be determined as yet), and open chromatin (regions of spaced-out histones, making them accessible to protein interactions). The interactive Genoverse genome browser (http://genoverse.org) has been developed by the DECIPHER team. Abbreviations: CTCF, CCCTC-binding factor; pHaplo, probability of haploinsufficiency.

Figure 5. DECIPHER shares cardiac case–control data collated by Cardiac VariantFX.

(a) For a given genomic variant, the allele frequency, allele count, and allele number observed in HCM, dilated cardiomyopathy, and healthy volunteer cohorts are displayed, assisting in the use of the PS4 criterion for variant interpretation (example displayed: https://www.deciphergenomics.org/sequence-variant/14-23429005-G-A/annotation/disease-cohorts/cardiac/hcm). (b) Summary cardiac case–control cohort data can be used to determine the confidence of gene–phenotype relationships and are displayed on gene tabs in DECIPHER (example displayed: https://www.deciphergenomics.org/gene/MYH7/overview/clinical-info). (c) Detailed case–control cohort metrics are available on a modal. Three metrics (excess in disease, etiological fraction, and odds ratio) are displayed for all variants, nontruncating variants, and truncating variants. Known mechanisms of disease are also described. In this example, distinct variants in MYH7 lead to HCM and dilated cardiomyopathy via opposing mechanisms—that is, activating variants cause HCM, and inactivating variants cause dilated cardiomyopathy. MYH7 is not known to be haploinsufficient, and null alleles are not associated with either condition. Abbreviation: HCM, hypertrophic cardiomyopathy.

Figure 6. DECIPHER brings together relevant data and contextualizes a patient’s variant with respect to available datasets in addition to providing annotations.

(a) The 2D protein browser provides a powerful genotypic overview (example displayed: https://www.deciphergenomics.org/gene/CDK13/overview/protein-genomic-info). DECIPHER and ClinVar variants are plotted individually as triangles, with filled triangles indicating a pathogenic/likely pathogenic annotation and open triangles indicating a benign/likely benign annotation or that the variant is not annotated. Due to the large number of gnomAD missense variants, these symbols are displayed as a histogram, with additional tracks displaying the location of homozygous missense, LOF, and homozygous LOF variants. Colors indicate the molecular consequence of variants predicted by Ensembl VEP; for example, red indicates a variant likely to have LOF consequences, such as a frameshift variant or splice acceptor variant, and yellow indicates a variant likely to have protein-changing consequences, such as a missense variant or in-frame deletion. (b) A 3D protein viewer is available that displays DECIPHER variants plotted on experimentally determined 3D structures (left) and 3D predicted structures (right). (c) ClinVar and ClinGen expert panel recommendations are shown (example displayed: https://www.deciphergenomics.org/sequence-variant/10-87952142-C-T/annotation/clinvar). (d) Functional data from the neXtProt knowledgebase are available (example displayed: https://www.deciphergenomics.org/sequence-variant/10-87952142-C-T/annotation/functional). Abbreviations: LOF, loss of function; gnomAD, Genome Aggregation Database; VEP, Variant Effect Predictor.

Figure 7. DECIPHER helps depositors share and compare patient phenotype and provides a summative assessment interface to help determine clinical fit.

(a) The matching patient variants interface provides a summary of patients with overlapping variants. (b) Disorder-specific centile charts are generated from quantitative data deposited to DECIPHER (example displayed: https://www.deciphergenomics.org/gene/CDK13/overview/clinical-info). (c) DECIPHER provides a summative assessment interface that provides a framework to determine whether a clinico-molecular diagnosis has been established.

Figure 8

DECIPHER facilitates empirical, hypothesis-driven research, such as the identification of a novel disorder associated with KCNK3, which is potentially treatable by pharmacological intervention. This example comprises six steps: (①) search on the DECIPHER website for patients with variants in KCNK3; (②) use DECIPHER’S matching patient variants interface to view overlapping phenotypes and identify a novel phenotype (central sleep apnea) not known to be associated with KCNK3 (③) use DECIPHER’S 2D and 3D protein viewers to observe the location of the missense variants; (④) contact the submitting centers through DECIPHER; (⑤) initiate a multicenter research effort involving clinicians, sleep specialists, genomic scientists, electrophysiologists, and pharmacologists; and (⑥) publish the results and add a citation to DECIPHER to share this information.