MITOMICS - PORTAIL DE TRANSPARENCE

Mitochondrial Disease database: An integrated multi-OMICS approach

Mitochondrial diseases are rare, clinically and genetically extremely heterogeneous, caused by a deficit of energy production via the mitochondria. Mitochondria are dependent on 2 genomes mitochondrial DNA (mtDNA) and nuclear DNA, and many pathogenic variants carried by these 2 genomes are responsible for mitochondrial diseases. There is a cross talk and regulatory mechanisms between both genomes, which are still poorly understood, involved in the control and maintenance of mitochondrial biogenesis. All of these mechanisms play an important role in the clinical and genetic heterogeneities presented by patients suffering from mitochondrial disease and are difficult to identify by "classical" high-throughput mtDNA, whole exome (WES) or genome (WGS) sequencing approaches. Mitomatcher is the first French database collecting genetic and clinical data for patients with mitochondrial diseases, implemented by the national mitochondrial laboratory network MitoDiag in conjunction with the reference centres (CARAMMEL and CALISSON) and Filnemus the rare disease network. Mitomatcher comprises 3 different modules: i) HPO related phenotypic data module, ii) genetic module currently containing mtDNA variants from more than 3000 patients with mitochondrial disease and iii) query and cross-reference module for the clinical-biological data. The MITOMICS project aims to better understand the molecular mechanisms responsible for the clinical-genetic heterogeneity of mitochondrial diseases. The integration of multi-Omics data (transcriptomics/proteomics/metabolomics) combined with clinical and genomic data (WES, WGS) in the Mitomatcher database should help to unravel the complexity of these diseases. Societal implications (ethics, juridics) of genomics research will be investigated and guidelines will be defined. The cross analysis of these data requires the development of in silico tools. Different approaches will be developed to (i) identify co-occurrences of mtDNA and/or nuclear DNA variants responsible for mitochondrial diseases in order to reveal new genotype/phenotype correlations, (ii) characterise the mitochondrial and nuclear crosstalk, and (iii) identify OMICS signatures specific to mitochondrial dysfunction. The study will be divided into 4 WPs starting with data collection and mitomatcher implementation (WP1) and from WP2-4 with the development of algorithms and data integration with increasing complexity with sequential data implementation starting with mtDNA variants alone, then combined with nuclear variants and finally multi-OMICs. Innovative machine learning, neural networks and artificial intelligence approaches will be developed such as Ruche a multi-layeR mUlti-omics maCHine learning intEgration tool or ABEILLE (ABerrant Expression Identification empLoying machine LEarning), an autoencoder-based method for the identification of aberrant gene expression from RNA-seq which will be applied to other OMICs. The identification of combinations of variants or affected signaling pathways from homogeneous groups of patients will be further verified by laboratory experiments. This project will also allow further development of in silico tools for the analysis of mtDNA variants such as Eklipse to detect mtDNA rearrangements. Mitomatcher database will be accessible, already following international standards (Fast Healthcare Interoperability Resources, HPO), interoperable with other national or international databases and reusable for the development of ancillary studies for these disorders. In the long term, the results obtained will allow the identification of new genotype/phenotype correlations and a better understanding of the pathophysiological mechanisms of mitochondrial diseases. The identification of specific signatures via an integrated multi-OMICs approach through Mitomics should also target specific pathways, thus enabling the development of new therapeutic strategies for mitochondrial diseases.

Charte_utilisation_mitomatcher_20211119.pdf