Extended Data Fig. 7: Relative transcript and protein abundance for dinoflagellate genes of interest between depths and across latitudes.

a, Relative gene expression is shown as log₂ transcripts per million (TPM; top), relative protein abundance as log₂ normalized spectral abundance factor (NSAF; bottom). Undetected transcripts and proteins are indicated in gray. The depth annotation bar indicates samples from the surface (<200 m, white), deep (>200 m, black) and 200 m (gray), and the site annotation bar shows whether samples were collected from the oligotrophic gyres (St. 1, 3, 9, 12) or the equatorial upwelling zone (St. 5, 6, 8). b, Comparison of average transcript and average protein abundance fold changes between euphotic and mesopelagic zones. Only shared KEGG genes (KOs) are shown that were detected in at least one metatranscriptome and one metaproteome, with proteorhodopin (PFam PF01036) and ISIP2a (KEGG gene pti:PHATDRAFT_54465) manually added. Values of zero were changed to a small value (0.1) to allow for fold changes estimates. The black line illustrates the linear relationship between protein and transcript fold changes. Genes in the top right (quadrant 1) represent transcripts/proteins abundant in euphotic waters, genes in the bottom left (quadrant 3) represent transcripts/proteins abundant in the mesopelagic (FBA = fructose biphosphate aldolase; GAP2 = glyceraldehyde-3-phosphate dehydrogenase (NAD(P)); ISIP2A = iron starvation induced protein 2 [phytotransferrin]; GAPDH = glyceraldehyde 3-phosphate dehydrogenase).