Abstract

The advancement of Next Generation Sequencing (NGS) generates a huge pool of raw sequencing data and genomic variants, while the diverse selection of variant annotation tools adds even more confusion to the mix. Choosing the right tools for clinical interpretation of genomic variants is still challenging due to the lack of comprehensive evaluation studies in this field. Here, we introduced a bias-free analysis approach to assess ten well-known variant annotation tools in terms of clinical interpretation. Our results revealed notable correlations of contemporary methods when applied to the Clin Var dataset. Moreover, allele frequency is still a strong predictor, emphasizing the importance of biological insight in the prediction of clinical interpretation. Our analysis and evaluation scripts are available for public use at https://github.com/nmtrang001Var_Annot_Eval.

Keywords: Training, Knowledge engineering, Sequential analysis, Reactive power, Annotations, Genomics, Market research