Back

Phylogenomics of Cypriniformes, the most diverse order of freshwater fishes: consensus, challenges and limitations

Sudasinghe, H.; Britz, R.; Matschiner, M.; Conway, K.; Tan, H. H.; Kottelat, M.; Raghavan, R.; Dahanukar, N.; Anoop, V. K.; Bohlen, J.; Howe, K.; Durbin, R.; Salzburger, W.; Peichel, C.; Rueber, L.

2026-01-14 evolutionary biology
10.64898/2026.01.14.699467 bioRxiv
Show abstract

Cypriniformes, the most species-rich order of freshwater fishes ([~]5,000 species), represents a key lineage for understanding vertebrate diversification in freshwater ecosystems. This clade includes several highly miniaturized and understudied lineages whose phylogenetic placements have long remained contentious. Here, we present the first phylogenomic analysis of Cypriniformes with complete family-level representation and broad genus-level coverage, encompassing 316 species comprising approximately 30% of all described genera. Our dataset integrates 257 newly assembled genomes with publicly available resources and analyzes multiple sets of genome-wide markers using both concatenation-based and coalescent-aware approaches. The general concordance among analytical frameworks indicates that the backbone topology of Cypriniformes is now established, allowing clear identification between well-supported clades and regions of persistent conflict. Our results strengthen the evolutionary relationships of several miniaturized lineages, while identifying recalcitrant relationships shaped by both biological processes and model artefacts that can yield superficially similar patterns of phylogenetic conflict. Our study substantially expands genomic representation and establishes a phylogenomic foundation for future comparative, developmental, and evolutionary research in this freshwater radiation.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.