Computational Biology
@ Comenius University in Bratislava

Martin Senderák, Tomáš Vinař. Seeing Through Hairballs: Simplifying Pangenome Graphs. In Information Technologies – Applications and Theory (ITAT), 4092 volume of CEUR Workshop Proceedings, pp. 326-333, 2025.

Download preprint: not available

Download from publisher: https://ceur-ws.org/Vol-4092/paper26.pdf

Related web page: not available

Bibliography entry: BibTeX

Abstract:

When building a pangenome graph of tens of individual bacterial genomes, one often encounters a problem
that the structure of the resulting graph is generally unwieldy. In particular, some vertices are likely to connect
sequences that appear in different contexts in different genomes. When visualizing such pangenomes, these
vertices pull together different unrelated parts of individual genomes, creating a difficult to analyze hairball.
In this paper, we explore this issue and propose a greedy heuristic algorithm to simplify pangenome graphs
by eliminating certain vertices and edges. This means that certain portions of the original genomes will no
longer be represented in the pangenome, and therefore it is important to minimize such graph modifications. We
demonstrate effectiveness of our approach on a set of 50 Escherichia coli isolates and we show that visualization
of such a simplified pangenome graph can unmask interesting features of the pangenome organization