Bioinformatics integration of microbiome and metabolomics data in the translational contextDeutsch, Leon (Avtor) Stres, Blaž (Mentor) bioinformaticsmetabolomicsmetagenomcisphysical inactivitydata integrationsystems biologymicrobiomeurinenuclear magnetic spectrometryHuman metabolism was studied in three different projects focusing on different levels of inactivity. Urine, liquor and serum metabolomics were used to assess the impact of nusinersen treatment in patients with spinal muscular atrophy. Urine samples were contrasted with samples from matching healthy cohort. In the PreTerm project, metabolomics (fecal and urine) and fecal microbial metagenomics were used to assess the differences between preterm and full-term born adults. In the X-Adapt project, urinary metabolomics was used to evaluate the 10-day training regime and the differences between trained and untrained individuals. In all projects, classification models based on different data sets were developed as a proof of principle and to foster their use in future studies or possibly in medical diagnostics. In addition, two workflows (GUMPP and MAGO tool) and a method to study physicochemical parameters (minimum pressure of piercing strength) were developed to study the microbiome and its environment, respectively. The final work resulted in the creation of the first Slovenian urine 1H-NMR database, which consists of 1200 urine metabolomes from different projects (PlanHab, Spinal Muscular Atrophy, X-Adapt, PreTerm, Healthy males and females) measured by 1H-NMR, outlining the baseline for future extensions. The entire database can be used to build machine-learning models for classification between different diseases or levels of physical activity at a national level.20222022-11-30 07:15:19Doktorsko delo/naloga142851VisID: 216931COBISS_ID: 132071939sl