Beyond sequence similarity: cross-phyla protein annotation by structural prediction and alignment

Background: Annotating protein function is a major goal in molecular biology, yet experimentally determined knowledge is often limited to a few model organisms. In non-model species, the sequence-based prediction of gene orthology can be used to infer function, however this approach loses predictive power with longer evolutionary distances. Here we propose a pipeline for the functional annotation of proteins using structural similarity, exploiting the fact that protein structures are directly linked to function and can be more conserved than protein sequences. Results: We propose a pipeline of openly available tools for the functional annotation of proteins via structural similarity (MorF: Morpholog Finder) and use it to annotate the complete proteome of a sponge. Sponges are highly relevant for inferring the early history of animals, yet their proteomes remain sparsely annotated. MorF accurately predicts the functions of proteins with known homology in >90% cases, and annotates an additional 50% of the proteome beyond standard sequence-based methods. Using this, we uncover new functions for sponge cell types, including extensive FGF, TGF and Ephrin signalling in sponge epithelia, and redox metabolism and control in myopeptidocytes. Notably, we also annotate genes specific to the enigmatic sponge mesocytes, proposing they function to digest cell walls. Conclusions: Our work demonstrates that structural similarity is a powerful approach that complements and extends sequence similarity searches to identify homologous proteins over long evolutionary distances. We anticipate this to be a powerful approach that boosts discovery in numerous -omics datasets, especially for non-model organisms.We upload revision.tar.gz, containing new files generated during the peer review process, as well as files that were challenging to find or recreate: - arabidopsis.m8: Foldseek alignment scores for the arabidopsis proteome - yeast.m8: Foldseek alignment scores for the yeast proteome - A_thaliana_annotations.tsv: EggNOG annotation of the A. thaliana proteome - S_cerevisiae_annotations.tsv: EggNOG annotation of the S. cerevisiae proteome - DEG_celltypes.tsv: differentially expressed genes for each celltype (Musser et al. 2021) - DEG_clades.tsv: differentially expressed genes for each clade (Musser et al. 2021) - DEG_clusters.tsv: differentially expressed genes for each cluster (Musser et al. 2021) - DEG_single_cell.tsv: all differentially expressed genes (Musser et al. 2021) - afdb_proteomes_species.tsv: UniProt ID to species map for all AFDB entries - cutoffs.tsv: top N hits bit score cutoffs for each Spongilla protein - filtered_afdb.tsv: filtered Foldseek output for AFDB self-alignment - self_score.tsv: unfiltered Foldseek output for AFDB self-alignment - slac_hmmer.emapper.annotations: output of EggNOG-mapper in HMMER mode for the Spongilla proteome - spongilla_basic.h5ad: main H5AD file used in the analysis; based on Seurat object from Musser et al. 2021 - spongilla_lut.tsv: annotation look-up table from Musser et al. 2021 - uni2egg_euk.Nov2018.tsv: (incomplete) mapping of UniProt ID to eukaryotic orthologous group. From EggNOG v5. - uniprotinfo.fasta: FASTA file of unique UniProt sequences found in the AFDB self-alignmen