Nombre de citations par an. Citations en double. Ajouter les coauteurs Coauteurs. Importer un PDF. Nouveaux articles de cet auteur. Nouvelles citations des articles de cet auteur.
|Published (Last):||20 October 2015|
|PDF File Size:||20.6 Mb|
|ePub File Size:||10.54 Mb|
|Price:||Free* [*Free Regsitration Required]|
Further analysis extended to all mutated genes in a research environment is a valuable strategy for improving diagnostic yields. We identified 17 new genes not related to human disease, implicated 22 non-OMIM disease-causing genes recently or previously rarely related to disease, and described 1 new phenotype associated with a known gene. Twenty-six candidate genes were identified and are waiting for future recurrence. This study demonstrates the power of such extended research reanalysis to increase scientific knowledge of rare diseases.
These novel findings can then be applied in the field of diagnostics. Over the last decade, next-generation sequencing has revolutionized the world of rare diseases. Two-thirds of these patients therefore remain without a molecular diagnosis after cES. This diagnostic yield is limited by the stringent criteria of the ACMG American College of Medical Genetics and Genomics , which recommends restricting variant interpretation to the genes responsible for human diseases [ 6 ] mostly reported in the OMIM database, using a routine practice of wide diagnostic laboratories.
The need for periodic reanalysis of ES data in undiagnosed patients therefore appears obvious. Such analyses could be enriched by the significant number of disease-causing genes published in the scientific literature and not referenced in the OMIM database. Moreover, information about variants has been updated in databases, such as OMIM or ClinVAR, because the reanalysis of ES data results in the reclassification of previously described variants.
These changes are mainly variants of uncertain significance reclassified as affect functions or benign variants [ 3 ]. The reanalysis and reinterpretation of ES data in a research setting are made possible by a wide variety of tools and databases, and the abundant knowledge available in the scientific literature.
A few studies that have used a range of currently available tools to reanalyze the cES data in the research environment have successfully increased diagnostic yield [ 3 , 8 , 16 ]. The diagnostic yield was dependent on the strategy and varied widely from study to study. Nambot et al. The combined results of these studies indicated that the OMIM database is slow to be updated and the drawbacks of restricting analyses to OMIM disease-causing genes were clearly demonstrated.
Another effective strategy for accelerating the identification of new disease-causing genes is the use of trio-based ES. Eldomery et al. These studies evidence the clear advantage of extending exploration to non-OMIM disease-causing genes, though further analyses will be essential to confirm the preliminary results [ 17 ]. Our findings supplement a previously published study in the diagnostic setting [ 8 ].
We discuss the potential gains and consequences of reanalyzing ES data in individual care and for scientific knowledge. Singleton-ES data were obtained from a cohort of unrelated patients, from families, referred to the Reference Center for Congenital Anomalies and Malformative Syndromes in Dijon France , or the Orphanomix units for genetics testing located in several hospitals in France.
These data were reanalyzed in a research laboratory between July and December Fig. The local ethics committee approved this study. Strategy of ES data analysis, databases, and tools used for variant interpretation in a research environment and the global results of this study. First, we filtered on the suspected mode of inheritance. Then, a large research, including literature and public databases associated with in silico scores, helping at the interpretation or indicates a possible link with the disease, allowed to select candidate genes.
Finally, accordingly with the parental segregation, the variant can be shared in international platforms for the purpose of the identification of additional affected cases to confirm or rule out the candidate gene. Each research analysis was done immediately after the cES results were obtained. Rapid deployment is facilitated by our local translational integrative organization that combines a unit focused on diagnostic innovation and a research team.
ES data are typically analyzed in two steps: 1 diagnostic interpretation restricted to disease-causing genes reported in the OMIM database, and 2 reanalysis in a research environment, including all detected variants.
In each stage, a multidisciplinary team is implicated in the interpretation of ES data and some experts are involved in the two steps.
We used all of the ES data initially produced for the singleton cES analysis [ 8 ]. All aligned reads underwent the same procedure: i duplicate paired-end reads were removed by Picard 2.
In the reanalysis for research purposes, we extended variant interpretation to genes not associated with human disease in the OMIM database. We first looked at a gene recently recognized in the literature as disease-causing but not listed in OMIM, and then we turned our attention to genes unknown to cause human diseases yet.
To help with the interpretation, we used public databases listing variants or genes previously reported in human diseases, animal models such as mouse, zebrafish, or rat , and gene expression data, and the impact of the variant in protein structure or function, protein interactions, or signaling pathways. We also used bioinformatics tools, including predictive scores of pathogenicity, conservation, or impact of missense and truncating variants Fig.
We performed a systematic review of the literature to identify isolated cases, recent description of new genes, or functional data. In this study, cES data have been analyzed by two experts, and all of the results and candidate genes were presented and discussed in multidisciplinary assembly.
Sequence data were analyzed with Mutation Surveyor v4. For each candidate variant in a candidate gene, unknown in rare diseases or known but with a new phenotypic presentation, we actively searched for additional similar patients to confirm the genotype—phenotype correlation Fig. Multiple strategies were used to improve data sharing: exchanging information by e-mail, presenting case-report results in international congresses, scrutinizing a large series of ES studies in the literature or in public databases i.
Eleven patients were born to reported consanguineous parents. The ethnic origin was European All individuals had received a negative or non-conclusive result after array-CGH and singleton cES analysis in the diagnostic setting. Of the individuals, were cases without a molecular diagnosis, who had been the focus of a previous study and 14 were individuals who have a candidate gene previously reported in the case of additional research strategy [ 8 ].
Of the 7. Three autosomal- dominant variants were confirmed by Sanger sequencing, but the parental segregation is unknown, because parental DNAs are not available Fig. Repartition of the mode of inheritance and type of variant in validated disease-causing genes a or in candidate genes b.
A match was found for 16 submitted genes. The number of recurrence to confirm the involvement of a gene in a disease is estimated to three unrelated cases with homozygous or compound heterozygous variants in autosomal recessive phenotypes, and five affected cases with heterozygous variants for autosomal-dominant phenotype [ 18 ].
In the absence of insufficient recurrence, the genes remain a candidate with a variant of uncertain significance.
The autosomal or X-linked sporadic variants appear to affect function, while half of autosomal-recessive or X-linked inherited variants remain variants of uncertain significance Figs. Data sharing resulted in national and international collaborations for 21 genes 27 patients , 15 of which were used for functional studies.
Our results have led to 15 scientific publications, and 12 papers are currently in progress, 6 of which are being led by our team [ 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 ]. Percentage of inconclusive with variants of uncertain significance or positive diagnosis by the mode of inheritance d or type of variants e.
Research reanalysis after negative singleton cES has demonstrated its ability to rapidly improve diagnostic yield and scientific knowledge. Analyzing singleton-ES data for research purposes provides a valuable opportunity to identify new disease-causing genes. Strategies combining research reanalysis and international data sharing foster national and international collaborations 21 collaborative projects and have improved scientific knowledge of rare disorders.
Our results have resulted in 15 papers in scientific publications [ 19 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 ] and 12 papers are currently in progress. Despite regular updates, the OMIM database remains incomplete and provides only limited resources for the diagnosis of rare diseases. Access to all available scientific literature is essential, because it allows researchers to identify genes previously published but not yet recognized in the OMIM database [ 11 ].
The TBR1 gene was first reported in patients with autism in , and more than ten unrelated patients were later reported in different large cohorts of autistic individuals. Our data sharing identified 20 additional individuals with ID and TBR1 variants, definitively establishing causality Nambot et al. Regular updates to the OMIM database encourage prospective diagnostic reanalysis.
This can lead to new diagnoses from recently identified genes that are progressively reported in the OMIM database, but the molecular diagnosis is generally delayed by at least 1 year [ 8 ].
Frequent literature reviews and prospective updates of bioinformatics pipelines would ensure the diagnosis of rare diseases linked to recently identified genes. Another fundamental challenge for variant interpretation is to provide a positive diagnosis for well-known genes, when the phenotype or genotype is atypical. Clinical heterogeneity is common in rare diseases, ranging from vast, overlapping clinical spectrums to completely different phenotypes [ 40 , 41 , 42 ].
In ultrarare diseases, knowledge is limited by the lack of individuals with causal variants in the same gene. Research analysis could therefore identify new phenotypes linked to well-known genes, but only recognized for only one phenotype with different clinical presentations in the OMIM database.
If there is no correlation with the OMIM phenotypic reference, reverse phenotyping limits the chances of delivering an initial diagnosis.
We submitted the case to the GeneMatcher platform and at an international congress in the hope of identifying additional patients, and thus recruited five additional patients with a Bohring—Opitz-like presentation. We finally expanded the clinical spectrum of KLHL7 autosomal-recessive variants by describing a syndrome with features overlapping cold-induced sweating syndrome 3 and Bohring—Opitz syndrome [ 23 ].
Variant interpretation in ES research analysis is mainly based on suspected inheritance, phenotypic, variant and functional databases, variant prediction scores, animal models, accessible literature, and data sharing Fig.
Alas, even with access to multiple data sources and the many tools designed to exploit them, many candidates remain Fig. The use of a singleton strategy for diagnosis has an economic advantage [ 8 ], but shows its limits when the analysis is extended to research.
Access to mutational parental segregation and detailed phenotyping would certainly be an advantage in this highly complex context. The absence of functional studies makes predicting the impact of missense variants difficult.
Algorithms have been developed to help with variant interpretation, but are not sufficient without biological validation Fig. In addition to the missense variants, a large portion of splice site variants remained candidates Fig. Our work shows that data sharing is essential for establishing human genotype—phenotype relationships and conclusively classifying variants. The Matchmaker Exchange Initiative offers a data-sharing platform that can be used to match patients all over the world, according to their phenotypic or genotypic features [ 47 ].
Data-sharing tools increase diagnostic yield and should be actively used by the scientific and medical communities. Because these genes are candidates for ultrarare phenotypes and the number of clinicians or scientists sharing genotypic data is relatively small, additional cases have not yet been identified. In time, the increasing amount of data shared on international platforms should provide additional chances to conclude [ 25 , 30 ]. In addition, the detection of variants remains, mainly depending of the sequencing technique, the exome capture kit, the exome coverage, and the bioinformatics pipeline used [ 49 , 50 , 51 ].
This study demonstrates how singleton-ES research reanalysis can efficiently and rapidly increase scientific knowledge in rare diseases by identifying new disease-causing genes, implicating recent known genes not reported in the OMIM database, or extending the phenotype or genotype spectrum of well-known genes.
The limits of the singleton strategy could be overcome with a second-step strategy based on trio ES. Complementary ES strategies increase the diagnostic ability of ES, and should be explored further and would be interesting to be quickly integrated after an initial diagnostic analysis in changing diagnostic laboratory practices in the strategy of exome analysis in the future. Enhanced utility of family-centered diagnostic exome sequencing with inheritance model-based analysis: results from unselected families with undiagnosed genetic conditions.
Page not found
Cours Execices Universite