Extended Data Fig. 8: Diversity and distribution of PRs and XRs with (G) and without (FW) fenestration among different prokaryotic phyla across four environments.

a, Maximum likelihood phylogenetic analysis of the PR-XR-NQ clade based on representative protein sequences. Characterized ion pumps are indicated with dots, terminal branches are colored by the corresponding phylum. Major clades with more than one representative are highlighted and labeled. The tree is outgroup rooted. b, Distribution of PRs and XRs with the canonical TM3 motif DTE among genomes assigned to different taxa, with (G) and without (FW) fenestration. The analysis is based on GEM genomes and the numbers are summarized per operational taxonomic unit (OTU). The colors are as in panel (a). c, Relative abundance of different families of the clade across four habitats based on the metagenomic data from IMG/M. Only families with a total relative abundance of >0.1% are shown. d, Predicted absorption maxima for PRs and XRs with the three most frequent residues at the fenestration position. Individual observation corresponds to an average absorption maximum predicted with the rhodopsin BLASSO model for sequences with the same 24 residues of the retinal binding pocket⁵⁶. The sequences from OM-RGC, IMG/M and GEM were pooled together. The size of the dots is proportional to the number of distinct rhodopsin domain sequences and the color approximates the predicted mean absorption spectra. Statistical differences between the groups were assessed with Dunn’s test with FDR correction. Significance levels are indicated with asterisks: *** – adjusted p-values < 0.001. Abbreviation of family names in (A) and (C): ACB – Archaea clade B, ESR – Exiguobacterium sibiricum rhodopsin, NQ – NQ sodium and chloride pumps, MACR – marine actinobacteria clade rhodopsins, PR – proteorhodopsins, TAT – TAT rhodopsins, XR – xanthorhodopsins, P1 – unnamed clade including QsActR, KrActR and related rhodopsins, P3 and P4 – currently unnamed clades.

Source data