This thesis presents a workflow for analysis of exome sequencing data aiming at identification of single nucleotide variants (SNVs) causing recessively inherited mitochondrial disease in children. Several variant selection criteria that are consistent with such group of genetic disorders are applied along the workflow in relation to mode of inheritance, allele frequency and the Finnish ancestry of the patients. These are combined with knowledge of nuclear-encoded mitochondrial proteins and prediction of pathogenic variants, narrowing down the total set of SNVs found in a patient to those most likely to be causative. Patient exomes are analysed individually (n=1 studies). The bioinformatic resources used for implementation include public and in-house databases of mitochondrial nuclear genes, human genetic variation and exome controls, as well as software tools for prediction of pathogenic SNVs and mitochondria-targeting proteins. Exome variant data from a cohort of 49 molecularly undiagnosed children were analysed through the workflow, leading to the identification of mitochondrial disease-causing SNVs located in nuclear genes for 10 of the patients. Therefore, a success rate of 20% was achieved. The workflow has been an important element in the use of exome sequencing as a new research tool at the Wartiovaara group of the Research Program for Molecular Neurology, Faculty of Medicine, University of Helsinki.
