Complete pangenome graph annotation with GrAnnoT and GrAPEx

Background: Pangenome variation graphs (PVG) are powerful structures for studying genetic diversity and minimizing bias in genomic research. Currently, it is not feasible to incorporate comprehensive annotation of the PVG. Our previous work introduced GrAnnoT, a tool for transferring annotations from a reference genome to a PVG and from a PVG to embedded genomes. However, GrAnnoT had limitations, only allowing one annotation to be transferred at a time and unable to accommodate multiple genomes.

Results: We have enhanced GrAnnoT with new features that now support annotations from multiple genomes, integrating them seamlessly into the graph. Additionally, we developed GrAPEx, a pipeline designed to detect, extract, and annotate non-annotated regions within the graph. We rigorously tested these tools to evaluate their performance.

Our benchmark results demonstrate that both GrAnnoT and GrAPEx are reliable and efficient. GrAnnoT now successfully incorporates multiple annotations, expanding the biological information available in the graph and offering valuable insights into gene distribution within the PVG. GrAPEx enables rapid annotation of the entire PVG without compromising the quality of the annotations.

Conclusion: By presenting these two tools, we have developed a comprehensive solution for thoroughly annotating the entire PVG and integrating biological annotation into pangenome graphs. This advancement effectively reduces the previous reference bias, where only a single annotation (typically the reference genome) was included in the graph. This new capability represents a significant addition to the pangenomic toolkit, greatly enhancing the accuracy and completeness of genomic studies.