Search

Exome sequencing used for molecular diagnosis of Mendelian disorders considerably increases the number of missense variants of unclear significance, whose pathogenicity can be assessed by a variety of prediction tools. As the performance of algorithms may vary according to the datasets, complementary specific resources are needed to improve variant interpretation. As a model, we were interested in the cystic fibrosis transmembrane conductance regulator gene (CFTR) causing cystic fibrosis, in which at least 40% of missense variants are reported. Cystic fibrosis missense analysis (CYSMA) is a new web server designed for online estimation of the pathological relevance of CFTR missense variants. CYSMA generates a set of computationally derived data, ranging from evolutionary conservation to functional observations from three-dimensional structures, provides all available allelic frequencies, clinical observations, and references for functional studies. Compared to software classically used in analysis pipelines on a dataset of 141 well-characterized missense variants, CYSMA was the most efficient tool to discriminate benign missense variants, with a specificity of 85%, and very good sensitivity of 89%. These results suggest that such integrative tools could be adapted to numbers of genes involved in Mendelian disorders to improve the interpretation of missense variants identified in the context of diagnosis., (© 2019 Wiley Periodicals, Inc.)

Most of the 2,000 variants identified in the CFTR (cystic fibrosis transmembrane regulator) gene are rare or private. Their interpretation is hampered by the lack of available data and resources, making patient care and genetic counseling challenging. We developed a patient-based database dedicated to the annotations of rare CFTR variants in the context of their cis- and trans-allelic combinations. Based on almost 30 years of experience of CFTR testing, CFTR-France (https://cftr.iurc.montp.inserm.fr/cftr) currently compiles 16,819 variant records from 4,615 individuals with cystic fibrosis (CF) or CFTR-RD (related disorders), fetuses with ultrasound bowel anomalies, newborns awaiting clinical diagnosis, and asymptomatic compound heterozygotes. For each of the 736 different variants reported in the database, patient characteristics and genetic information (other variations in cis or in trans) have been thoroughly checked by a dedicated curator. Combining updated clinical, epidemiological, in silico, or in vitro functional data helps to the interpretation of unclassified and the reassessment of misclassified variants. This comprehensive CFTR database is now an invaluable tool for diagnostic laboratories gathering information on rare variants, especially in the context of genetic counseling, prenatal and preimplantation genetic diagnosis. CFTR-France is thus highly complementary to the international database CFTR2 focused so far on the most common CF-causing alleles., (© 2017 Wiley Periodicals, Inc.)

By family-based screening, first Fuchs and then many other authors showed that mutations in THAP1 (THAP [thanatos-associated protein] domain-containing, apoptosis-associated protein 1) account for a substantial proportion of familial, early-onset, nonfocal, primary dystonia cases (DYT6 dystonia). THAP1 is the first transcriptional factor involved in primary dystonia and the hypothesis of a transcriptional deregulation, which was primarily proposed for the X-linked dystonia-parkinsonism (DYT3 dystonia), provided thus a new way to investigate the possible mechanism underlying the development of dystonic movements. Currently, 56 families present with a THAP1 mutation; however, no genotype/phenotype relationship has been found. Therefore, we carried out a systematic review of the literature on the THAP1 gene to colligate all reported patients with a specific THAP1 mutation and the associated clinical signs in order to describe the broad phenotypic continuum of this disorder. To facilitate the comparison of the identified mutations, we created a Locus-Specific Database (UMD-THAP1 LSDB) available at http://www.umd.be/THAP1/. Currently, the database lists 56 probands and 43 relatives with the associated clinical phenotype when available. The identification of a larger number of THAP1 mutations and collection of high-quality clinical information for each described mutation through international collaborative effort will help investigating the structure-function and genotype-phenotype correlations in DYT6 dystonia., (© 2011 Wiley Periodicals, Inc.)

New technologies, which constantly become available for mutation detection and gene analysis, have contributed to an exponential rate of discovery of disease genes and variation in the human genome. The task of collecting and documenting this enormous amount of data in genetic databases represents a major challenge for the future of biological and medical science. The Locus Specific Databases (LSDBs) are so far the most efficient mutation databases. This review presents the main types of databases available for the analysis of mutations responsible for genetic disorders, as well as open perspectives for new therapeutic research or challenges for future medicine. Accurate and exhaustive collection of variations in human genomes will be crucial for research and personalized delivery of healthcare., (Copyright © 2009 Elsevier Masson SAS. All rights reserved.)

With the increasing knowledge of cystic fibrosis (CF) and CFTR-related diseases (CFTR-RD), the number of sequence variations in the CFTR gene is constantly raising. CF and particularly CFTR-RD provide a particular challenge because of many unclassified variants and identical genotypes associated with different phenotypes. Using the Universal Mutation Database (UMD) software we have constructed UMD-CFTR (freely available at the URL: http://www.umd.be/CFTR/), the first comprehensive relational CFTR database that allows an in-depth analysis and annotation of all variations identified in individuals whose CFTR genes have been analyzed extensively. The system has been tested on the molecular data from 757 patients (540 CF and 217 CBAVD) including disease-causing, unclassified, and nonpathogenic alterations (301 different sequence variations) representing 3,973 entries. Tools are provided to assess the pathogenicity of mutations. UMD-CFTR also offers a number of query tools and graphical views providing instant access to the list of mutations, their frequencies, positions and predicted consequences, or correlations between genotypes, haplotypes, and phenotypes. UMD-CFTR offers a way to compile not only disease-causing genotypes but also haplotypes. It will help the CFTR scientific and medical communities to improve sequence variation interpretation, evaluate the putative influence of haplotypes on mutations, and correlate molecular data with phenotypes., (Copyright 2010 Wiley-Liss, Inc.)

The collection of genetic variants that cause inherited disease (causative mutation) has occurred for decades albeit in an ad hoc way, for research and clinical purposes. More recently, the access to collections of mutations causing specific diseases has become essential for appropriate genetic health care. Because information has accumulated, it has become apparent that there are many gaps in our ability to correctly annotate all the changes that are being identified at ever increasing rates. The Human Variome Project (www.humanvariomeproject.org) was initiated to facilitate integrated and systematic collection and access to this data. This manuscript discusses how collection of such data may be facilitated through new software and strategies in the clinical genetics and diagnostic laboratory communities.

Approximately half of gene lesions responsible for human inherited diseases are due to an amino acid substitution, showing that this mutational mechanism plays a large role in diseases. Distinguishing neutral sequence variations from those responsible for the phenotype is of major interest in human genetics. Because in vitro validation of mutations is not always possible in diagnostic settings, indirect arguments must be accumulated to define whether a missense variation is causative. To further differentiate neutral variants from pathogenic nucleotide substitutions, we developed a new tool, UMD-Predictor. This tool provides a combinatorial approach that associates the following data: localization within the protein, conservation, biochemical properties of the mutant and wild-type residues, and the potential impact of the variation on mRNA. To evaluate this new tool, we compared it to the SIFT, PolyPhen, and SNAP software, the BLOSUM62 and Yu's Biochemical Matrices. All tools were evaluated using variations from well-validated datasets extracted from four UMD-LSDB databases (UMD-FBN1, UMD-FBN2, UMD-TGFBR1, and UMD-TGFBR2) that contain all published mutations of the corresponding genes, that is, 1,945 mutations, among which 796 different substitutions corresponding to missense mutations. Our results show that the UMD-Predictor algorithm is the most efficient tool to predict pathogenic mutations in this context with a positive predictive value of 99.4%, a sensitivity of 95.4%, and a specificity of 92.2%. It can thus enhance the interpretation of variations in these genes, and could easily be applied to any other disease gene through the freely available UMD generic software (http://www.umd.be).

UMD-DMD France is a knowledgebase developed through a multicenter academic effort to provide an up-to-date resource of curated information covering all identified mutations in patients with a dystrophinopathy. The current release includes 2,411 entries consisting in 2,084 independent mutational events identified in 2,046 male patients and 38 expressing females, which corresponds to an estimated number of 39 people per million with a genetic diagnosis of dystrophinopathy in France. Mutations consist in 1,404 large deletions, 215 large duplications, and 465 small rearrangements, of which 39.8% are nonsense mutations. The reading frame rule holds true for 96% of the DMD patients and 93% of the BMD patients. Quality control relies on the curation by four experts for the DMD gene and related diseases. Data on dystrophin and RNA analysis, phenotypic groups, and transmission are also available. About 24% of the mutations are de novo events. This national centralized resource will contribute to a greater understanding of prevalence of dystrophinopathies in France, and in particular, of the true frequency of BMD, which was found to be almost half (43%) that of DMD. UMD-DMD is a searchable anonymous database that includes numerous newly developed tools, which can benefit to all the scientific community interested in dystrophinopathies. Dedicated functions for genotype-based therapies allowed the prediction of a new multiexon skipping (del 45-53) potentially applicable to 53% of the deleted DMD patients. Finally, such a national database will prove to be useful to implement the international global DMD patients' registries under development.

Congenital contractural arachnodactyly (CCA) is an extremely rare disease, due to mutations in the FBN2 gene encoding fibrillin-2. Another member of the fibrillin family, the FBN1 gene, is involved in a broad phenotypic continuum of connective-tissue disorders including Marfan syndrome. Identifying not only what is in common but also what differentiates these two proteins should enable us to better comprehend their respective functions and better understand the multitude of diseases in which these two genes are involved. In 1995 we created a locus-specific database (LSDB) for FBN1 mutations with the Universal Mutation Database (UMD) tool. To facilitate comparison of identified mutations in these two genes and search for specific functional areas, we created an LSDB for the FBN2 gene: the UMD-FBN2 database. This database lists 26 published and six newly identified mutations that mainly comprise missense and splice-site mutations. Although the number of described FBN2 mutations was low, the frequency of joint dislocation was significantly higher with missense mutations when compared to splice site mutations., ((c) 2008 Wiley-Liss, Inc.)

Using the Universal Mutation Database (UMD) software, we have constructed "UMD-USHbases", a set of relational databases of nucleotide variations for seven genes involved in Usher syndrome (MYO7A, CDH23, PCDH15, USH1C, USH1G, USH3A and USH2A). Mutations in the Usher syndrome type I causing genes are also recorded in non-syndromic hearing loss cases and mutations in USH2A in non-syndromic retinitis pigmentosa. Usher syndrome provides a particular challenge for molecular diagnostics because of the clinical and molecular heterogeneity. As many mutations are missense changes, and all the genes also contain apparently non-pathogenic polymorphisms, well-curated databases are crucial for accurate interpretation of pathogenicity. Tools are provided to assess the pathogenicity of mutations, including conservation of amino acids and analysis of splice-sites. Reference amino acid alignments are provided. Apparently non-pathogenic variants in patients with Usher syndrome, at both the nucleotide and amino acid level, are included. The UMD-USHbases currently contain more than 2,830 entries including disease causing mutations, unclassified variants or non-pathogenic polymorphisms identified in over 938 patients. In addition to data collected from 89 publications, 15 novel mutations identified in our laboratory are recorded in MYO7A (6), CDH23 (8), or PCDH15 (1) genes. Information is given on the relative involvement of the seven genes, the number and distribution of variants in each gene. UMD-USHbases give access to a software package that provides specific routines and optimized multicriteria research and sorting tools. These databases should assist clinicians and geneticists seeking information about mutations responsible for Usher syndrome.

The implication of mutations in the TGFBR2 gene, known to be involved in cancers, in Marfan syndrome (MFS) and later in Loeys-Dietz syndrome (LDS) and Familial Thoracic Aortic Aneurysms and Dissections (TAAD2) gives a new example of the complexity of one gene involved in multiple diseases. To date, known TGFBR2 mutations are not disease-specific and many mutations have to be accumulated before genotype-phenotype relationships emerge. To facilitate mutational analysis of the TGFBR2 gene, a locus-specific database has been set up with the Universal Mutation Database (UMD) software. The version of the computerized database contains 85 entries. A total of 12 mutations are reported to be involved in MFS, six in incomplete MFS, 30 in LDS type I, 10 in LDS type II, seven in TAAD2, and 20 in various cancers. The database is accessible online at http://www.umd.be (last accessed: 3 July 2007)., ((c) 2007 Wiley-Liss, Inc.)

Mutation of the p53 gene is the most frequent genetic alteration found in human cancer, but it is also the most frequently reported with more than 22,000 mutations published in 2,000 papers. In 1991, we developed a database and software to handle and analyze all this information. The database has been widely used for clinical analysis and molecular epidemiology. We have expanded the scope of the database by integrating structural, phylogenetic and biological information on wild-type (wt) and mutant TP53. Integration of the TP53 mutant activity database provides unique information that will be useful to both clinicians and scientists. All of this information is available from a new website (www.umd.be:2072/) that will generate a detailed informative page for every TP53 mutant in the database. New tools to check TP53 mutations and minimize errors found in the literature are also available., (2005 Wiley-Liss, Inc.)

With the completion of the Human Genome Project, our vision of human genetic diseases has changed. The cloning of new disease-causing genes can now be performed in silico, and thousands of mutations are being identified in diagnostic and research laboratories yearly. Knowledge about these mutations and their association with clinical and biological data is essential for clinicians, geneticists, and researchers. To collect and analyze these data, we developed a generic software called Universal Mutation Databases (UMD) to create locus-specific databases. Here we report the new release (September 2004) of this freely available tool (www.umd.be), which allows the creation of LSDBs for virtually any gene and includes a large set of new analysis tools. We have implemented new features to integrate noncoding sequences, clinical data, pictures, monoclonal antibodies, and polymorphic markers (SNPs). Today the UMD retains all specifically designed tools to analyze mutations at the molecular level, as well as new sets of routines to search for genotype-phenotype correlations. We also created specific tools for infrequent mutations such as gross deletions and duplications, and deep intronic mutations. A large set of dedicated tools are now available for intronic mutations, including methods to calculate the consensus values (CVs) of potential splice sites and to search for exonic splicing enhancer (ESE) motifs. In addition, we have created specific routines to help researchers design new therapeutic strategies, such as exon skipping, aminoglycoside read-through of stop codons, or monoclonal antibody selection and epitope scanning for gene therapy.

Mutation databases of human genes are assuming an increasing importance in all areas of health care. In addition, more and more experts in the mutations and diseases of particular genes are curating published and unpublished mutations in locus-specific databases (LSDB). These databases contain such extensive information that they have become known as knowledge bases. We analyzed these databases and their content between June 21, 2001, and July 18, 2001. We were able to access 94 independent websites devoted to the documentation of mutation containing 262 LSDBs for study. We analyzed one LSDB from each of these websites (i.e., 94 LSDBs) for the presence or absence of 80 content criteria, as generally each gene in a multigene website documented the same criteria. No criterion studied gave unanimous agreement in every database. Twenty-two genes were represented by more than one LSDB. The number of mutations recorded, excluding p53, was 23,822 with 1518 polymorphisms. Fifty-four percent of the LSDBs studied were easy to use and 11% hard to follow; 73% of the databases were displayed through HTML. Three databases were found that were given a high score for ease of use and wealth of content. Thus, the study provided a strong case for uniformity of data to make the content maximally useful. In this direction, a hypothetical content for an ideal LSDB was derived. We also derived a community structure that would enhance the chances of mutation capture rather than being left unpublished in a patient's report. We hope the interested community and granting bodies will assist in achieving the vision of a public system that collects and displays all variants discovered.