Olfaction, the sense of smell, is a versatile and sensitive mechanism for detecting and discriminating thousands of volatile odorants. Olfactory recognition is mediated by large repertoires of olfactory receptors (ORs), which activate a G-protein-mediated transduction cascade, located in the cilia of olfactory sensory neurons [1, 2]. The human OR repertoire has 851 loci, encompassing 78 genomic clusters and 57 singleton loci, residing on all but two human chromosomes [3–6]. Each sensory cell expresses a single allele of a single OR locus, thus transmitting a molecularly defined signal to the brain [7–10]. A single OR gene may recognize more than a single odorant molecule [11–15]. A widely accepted working hypothesis is that allelic variants of OR genes may harbor different functional characteristics and hence, may generate different odorant sensitivity phenotypes in different members of the human population [16–18].
Human ORs encompass a high number of pseudogenes, whereby more than 50% of the loci annotated as nonfunctional due to frame-disrupting mutations [3, 5, 6, 19]. Primates are less dependent than mouse and dog on olfactory cues, which appears to have resulted in a gradual gene loss process along this lineage [20–22]. Similar OR repertoire diminutions have been reported in other mammals . In higher apes, the gene loss has remarkably accelerated in humans . Such diminution of the functional OR repertoire in humans is an ongoing evolutionary process, as demonstrated by the past identification of OR genes that segregate between intact and pseudogene forms [25, 26], and by more recent surveys showing an enrichment of loss-of-function OR alleles [27, 28]. It was shown that every human individual is characterized by a different combination of such segregating pseudogenes (SPGs), constituting a pronounced genotypic diversity in the population, including ethnogeographic differences . More recently, using a high-resolution microarray applied to 20 individuals , and a read-depth-based Copy Number Variation (CNV) genotyping algorithm , we showed a wide range of copy-number values across individuals, ranging from zero to nine copies. These results are in-line with other surveys which found a significant enrichment of ORs in CNV regions [31, 32]. CNVs involving deletions (copy numbers of 0 or 1) were shown to affect 56 intact OR loci, 14% of the human OR gene repertoire .
Cell-surface receptors are often characterized by several haplotypic alleles in the population, sometimes with different functional properties. A prominent example is the group of the major histocompatibility proteins with varying specificities towards antigenic peptides [33, 34]. Other examples include the taste receptor TAS38, underlying responsiveness to the bitter compound phenylthiocarbamide (PTC) [35, 36], the melanocortin 1 receptor (MC1R), affecting human skin and hair pigmentation , and the green opsin OPN1MW, mediating red-green color vision discrimination . Likewise, in the olfactory system, two protein haplotypes of the olfactory receptor OR7D4 were shown to manifest large difference in sensing the steroid odorant androstenone [39, 40].
Some missense haplotypic alleles can be nonfunctional, due to a substitution of key amino acids governing protein folding or interaction with signal transduction components. A continuous spectrum of functionality among missense haplotypes may be quantified by algorithms such SIFT  or PolyPhen . An analogous algorithm, Classifier for Olfactory Receptor Pseudogenes (CORP) , was previously used to identify 30 SNP variations for which one of the alleles is likely inactive , with a broader estimate of as many as 135 functionally inactive missense alleles in the reference genome .
Here, we performed scrutiny of publicly available data to create a comprehensive catalog of genetic variability in the human OR repertoire. This includes a compendium of all available missense haplotypes of OR proteins and a dramatically expanded list of OR segregating pseudogenes. Our work creates a framework for understanding the evolution and function of OR genes, and a necessary infrastructure for genotype-phenotype association studies for smell deficits. It further highlights the utility of the olfactory system as a model for personalized gene repertoires.