Yeast is a diverse group of organisms that has wide range of applications in food industry and human health. Due to complexity of the microbiome and errors in sequencing data, metagenomic analysis can be challenging. In this thesis work, factors influencing metagenomic analysis pipelines for yeast microbiomes were examined using modern in-silico methods of developing controlled simulated microbiomes. Utilizing Gargammel software, ancient sourdough and faecal metagenomes were simulated with Saccharomyces boulardii as the reference genome. The simulation process introduced nucleotide changes (deamination) and incorporated sequencing errors to mimic DNA damage. The generated FASTQ sequences for the ancient coprolite and sourdough metagenomes were subsequently analysed. The analysis showed a significant discrepancy between the nominal and actual coverage of various yeast species in the sourdough sample. These discrepancies suggest potential errors or inconsistencies in the mapping process. The analysis further continued with observing the impact of including additional yeast species in phylogenetic network analysis. Notably, there was a major difference in results between mitochondrial and nuclear networks. The additional fungal species provided more reference points, resulting in a varied representation of the evolutionary relationships between the yeasts, but only in mitochondrial network.