Fig 1.
Phylogenetic comparison of GII.4 variants.
(A) AA tree. (B) near-full-genome tree. (C) nucleotide tree. In the AA tree, variants form distinct clusters, and their phylogenetic relationships are consistent with their emergence over time. In contrast, the topologies of the full-length and nucleotide trees differ markedly from that of the AA tree. Trees were rooted with GII.17, GII.3, and GII.20 as outgroups.
Table 1.
Predicted low-confidence regions (pLDDT < 70) across norovirus genogroups/genotypes.
Fig 2.
Confidence assessment of the predicted structures.
(A) Distribution of pLDDT values for all predicted structures. The pLDDT values across all predicted structures had an overall mean of 93.9. (B) Residue-level pLDDT values for the major genotypes/groups. Most regions exhibited high confidence, with only a few loop regions showing lower values. (C) Structural comparison between predicted models of major genotypes and experimentally determined structures. The overall structures are highly consistent, though some local deviations were observed in loop regions. The experimentally determined structures were retrieved from the Protein Data Bank (https://www.rcsb.org) with accession numbers 5IYR (GII.4), 7ER0 (GI.3), 6IR5 (GII.3), and 5F4J (GII.17). (D) Distribution of predicted TM-scores (pTM). All predicted structures exhibited pTM above 0.85.
Fig 3.
Structurally aligned P domain phylogenies.
(A) 3Di_tree; (B) AA_tree; (C) AA + 3Di joint_tree. All three trees are consistent with the established viral classification but reveal discrepancies in specific phylogenetic relationships. Genotypes exhibiting inconsistent placements across the trees are highlighted in red. Sequences that could not be classified are labeled as “unknown” in each tree.
Fig 4.
Mean RMSD among genogroups and major genotypes.
The structural divergence between different genogroups is overall greater than that observed between different genotypes within the same genogroup. The highest inter-genogroup mean RMSD was observed between GI and GII, while the lowest was between GII and GIX. Among the major circulating genotypes, the smallest mean RMSD was found between GII.4 and GII.3, and the largest between GII.4 and GII.17.
Fig 5.
Residue conservation analysis.
(A-G) The per-residue RMSD contribution. Genotype GII.4, GII.17, GII.2, GII.3; GII.6, GI.6 and GI.3. GII.2 and GI.6 exhibited the highest structural conservation, followed by GII.17 and GII.4. (H) Informative AA substitutions. Residues mapping to the major antigenic sites of GII.4 showed substantial diversity.
Fig 6.
Structural and selective pressure analysis of the P region of GII.4.
(A) residue conservation mapping. Localized regions of high structural variability generally corresponded to known antigenic epitopes. (B) Spatial distribution of the major antigenic sites. (C) Distribution of positively selected sites. Among the 32 positively selected sites identified in the P domain, 19 are located in the P2 subdomain, with 15 falling within the major antigenic epitopes. “Purple circle” marked sites were identified by two methods, and “red circle” marked sites were identified by all three methods, and the rest unmarked sites were identified by only MEME method.
Fig 7.
Coevolving residue pairs and potential structural interactions in GII.4 P domain.
Among 31 coevolving pairs identified, 17 involve antigenic sites, predominantly in epitope A (12 pairs). Four pairs are in close spatial proximity (<8 Å), suggesting direct structural compensation. Four of the seven hub sites located in the major epitopes.