Conserved pattern-based classification of human odorant receptor multigene family

Conserved protein-coding sequences are critical for maintaining protein function across species. Odorant receptors (ORs), a large poorly understood multigene family responsible for odor detection, lack comprehensive classification methods that reflect their functional diversity. In this study, we propose a new approach called conserved motif-based classification (CMC) for classifying ORs based on amino acid sequence similarities within conserved motifs. Specifically, we focused on three well-conserved motifs: MAYDRYVAIC in TM3, KAFSTCASH in TM6, and PMLNPFIY in TM7. Using an unsupervised clustering technique, we classified human ORs (hORs) into two main clusters with six sub-clusters. CMC partly reflects previously identified subfamilies, revealing altered residue positions among the sub-clusters. These altered positions interacted with specific residues within or adjacent to the transmembrane domain, suggesting functional implications. Furthermore, we found that the CMC correlated with both ligand responses and ectopic expression patterns, highlighting its relevance to OR function. This conserved motif-based classification will help in understanding the functions and features that are not understood by classification based solely on entire amino acid sequence similarity.