Limitations of variant consequence predictors

Thanks to numerous large-scale sequencing projects, the number of discovered genomic variants is increasing. The key step in analyzing the variant data is the functional annotation, since it helps researchers and clinicians to categorize, filter and prioritize the variants for further research. This thesis discusses five commonly-used variant consequence predictors, offers advice on how to use them and briefly goes through the algorithms they employ. Moreover, various data formats as well as the human reference genome and different genome annotations are described in the thesis. The correctness of the reference is of great importance as all the predictors rely on it. This thesis highlights some situations in which the results given by different predictors can vary. All the tests were made using the Ensembl gene annotation (release 92) and the GRCh38 reference assembly