The Complete Human Olfactory Subgenome

Abstract

Olfactory receptors likely constitute the largest gene superfamily in the vertebrate genome. Here we present the nearly complete human olfactory subgenome elucidated by mining the genome draft with gene discovery algorithms. Over 900 olfactory receptor genes and pseudogenes (ORs) were identified, two-thirds of which were not annotated previously. The number of extrapolated ORs is in good agreement with previous theoretical predictions. The sequence of at least 63% of the ORs is disrupted by what appears to be a random process of pseudogene formation. ORs constitute 17 gene families, 4 of which contain more than 100 members each. “Fish-like” Class I ORs, previously considered a relic in higher tetrapods, constitute as much as 10% of the human repertoire, all in one large cluster on chromosome 11. Their lower pseudogene fraction suggests a functional significance. ORs are disposed on all human chromosomes except 20 and Y, and nearly 80% are found in clusters of 6–138 genes. A novel comparative cluster analysis was used to trace the evolutionary path that may have led to OR proliferation and diversification throughout the genome. The results of this analysis suggest the following genome expansion history: first, the generation of a “tetrapod-specific” Class II OR cluster on chromosome 11 by local duplication, then a single-step duplication of this cluster to chromosome 1, and finally an avalanche of duplication events out of chromosome 1 to most other chromosomes. The results of the data mining and characterization of ORs can be accessed at the Human Olfactory Receptor Data Exploratorium Web site (http://bioinfo.weizmann.ac.il/HORDE).

DEFOG: A Practical Scheme for Deciphering Families of Genes

“…We experimented with primers with even higher degeneracies, but they usually gave very poor yield (data not shown). In addition to the higher coverage, primers with higher degeneracy also allow higher sequencing efficacy: for most primer pairs with very high degeneracy (10 9 or more), the ratio between the number of different genes found and the number of clones sequenced was 0.79 to 0.93, whereas for primers with lower degeneracy, this ratio was 0.57 to 0.79.The pseudogene proportion in the sequenced set was about 55%, which is in good agreement with the ratio shown for the entire OR repertoire [9,10]. This suggests that the primer selection method did not create a bias of genes versus pseudogenes, which could result from higher sequence conservation in intact genes.…”

“…To further understand the reason for the two impure clusters, we compared the mean fingerprints of the two major classes in each of these clusters. The comparison gives a partial explanation to the mixed clusters: the mean fingerprints of the two classes in each cluster are very similar (their dot product was 0.7 out of a maximum of 1), which is in part due to sequence similarity (data not shown).The DEFOG experiment revealed 300 OR genes and pseudogenes, which is over 40% of the recently published full human OR repertoire [9,10]. Of these only 69 were encompassed in the training set of 127 ORs used for the primers design.…”

“…So far, about 3000 OR genes and pseudogenes are known in 24 vertebrate species [7][8][9][10][11][12]. They are divided into 32 distinct families based on phylogenetic analysis [8].Roughly 900 OR coding sequences were found in the first draft of the human genome [9,10]. As predicted [7,13], between 53% and 63% of them have frame disruptions and are therefore considered as pseudogenes.…”

“…2) shows that there was no cloning bias towards a certain closely related group of sequences. Low representation of families 51, 52, 55, and 56 is probably due to the low sequence similarity of these genes to the genes of the other OR families [9]. …”

“…characterized in rat a decade ago [2] and have since been detected in many vertebrate species [reviewed in 6]. So far, about 3000 OR genes and pseudogenes are known in 24 vertebrate species [7][8][9][10][11][12]. They are divided into 32 distinct families based on phylogenetic analysis [8].…”