Optimization of statistical and bioinformatic methods for the analysis of next generation sequencing data for rare disease diagnosis

Abstract

The main focus of this thesis, presented as a compendium of research articles, is the optimization of the analysis of Next Generation Sequencing data in order to facilitate the diagnosis of rare diseases. For this goal, we present an appropach to prioritize single nucleotide variants and small insertions and deletions, not only in terms of their type and genomic position, but also in terms of the mutational tolerance of the gene encompassing them. We also evaluate the strengths and weakness of the currently published copy number variation (CNV) detection tools, and develop a methodology to create sinthetic samples with artificial CNVs to test them. Finally, we present a novel CNV-detection program, optimized for gene panel assays.