Pangenomes, which capture the genetic diversity of populations more comprehensively than traditional linear genomes, are foundational to understanding genetic variation in species. While calculating statistical metrics for linear genomes can often be achieved with basic scripts, analyzing graph-based genomes requires efficient algorithms due to their complexity.
To address this challenge, a research team led by Jianyu Zhou published their novel tool GFAKaleidos on 15 April 2026 in Frontiers of Computer Science co-published by Higher Education Press and Springer Nature.
GFAKaleidos integrates three essential graph models—directed graphs, bidirected graphs, and biedged graphs—into one unified framework. This tri-model approach provides users with a “kaleidoscopic” perspective on the structural statistics of pangenome graphs, enabling comprehensive analysis of vertices, edges, and subgraphs, including features such as connected components, bubbles, loops, and cycles.
Compared with existing tools such as ODGI and gfatools, GFAKaleidos distinguishes itself through its comprehensive statistical outputs, intuitive visualizations, and flexible deployment options. Its user-friendly design caters to both beginners and experts, lowering the barrier to entry for high-level pangenome analysis and promoting broader adoption across the genomics research community.
The tool is freely available at https://combiopt.nankai.edu.cn/gfakaleidos and is being continuously developed to support more complex graph formats and to integrate with downstream functional genomic analyses.
Frontiers of Computer Science
Experimental study
Not applicable
GFAKaleidos: a tool for computing and comparing pangenome graph statistics
15-Apr-2026