Authors

Discussion

Cophylogeny attempts to estimate the past relationships between ecologically linked groups of organisms, such as hosts and parasites, based on comparison of their phylogenetic relationships. A central assumption is that congruence between the phylogenies of hosts and their parasites indicates a shared evolutionary history. However, perfect phylogenetic congruence is rarely observed, because several eco-evolutionary processes promoting incongruence can act concurrently. This often leads to situations that can be described as coevolutionary Gordian knots. Whereas event-based approaches focus on disentangling the knots, we propose here an Alexandrian solution: Random Tanglegram Partitions (Random TaPas). Random TaPas is particularly intended for analysis and visualization of large, entangled cophylogenetic settings. Given a tanglegram, consisting of two phylogenies, representing each the relationships among hosts and parasites, and the information of associations them across both phylogenies, Random TaPas resorts to recursive random partitions to establish heuristically (1) to which extent the relationships observed are due to cospeciation, (2) which individual host-parasite associations taxa contribute most to cospeciation and (3) whether cospeciation events are evenly distributed or concentrated in parts of the phylogenies. A desirable property of Random TaPas is that it can explicitly take into account phylogenetic uncertainty in the analysis. We assessed the performance of Random TaPas with simulated coevolutionary histories built using a varying number of common coevolutionary events (cospeciation, lineage duplication, failure-to-diverge, loss, spreading and host switching) and demonstrate its applicability with a real dataset.