Third-generation sequencing technologies, especially nanopores, present the possibility of fast sequencing DNA and obtaining long reads. Their downsides are high error rates. In this thesis, we present a bioinformatics pipeline for processing microsatellite sequences obtained using third-generation sequencing. For testing, we used brown bear (Ursus arctos) microsatellite sequences obtained from non-invasive genetic samples. They were sequenced on the Illumina platform. In these sequences, we simulated substitutions, insertions, and deletions with various combinations and different total error rates. Aside from the previously used 8 bp DNA tags for sample marking, we also tested longer 12 and 16 bp tags. Our bioinformatics pipeline was effective when dealing with substitutions only. It was ineffective when all three error types were simulated. Nonetheless, we found that the currently used 8 bp tags are not useful at high error rates, especially when dealing with all three error types. We also found issues with the primer search, and, consequently, identification of loci that are marked by the primers. We identified weak points in the pipeline and thus suggest possible solutions. The presented bioinformatics pipeline should therefore provide a useful basis for further work.
|
