Gene and whole genome duplications have been pivotal in vertebrate genome evolution [24]. Therefore, it is pertinent to explore the frequency of duplicated and unduplicated genes in different vertebrate species to understand the evolution of the vertebrate genome. Thus, we chose a comprehensive set of species from major vertebrate clades. As expected, duplicated genes comprised the predominant part of vertebrate protein-coding genomes, with median values of 79% in mammals, 76% in birds, 78% in reptiles, and 84% in fish (Fig. 1A,D). One reason fish species exhibited the highest duplicated genes is that fish undergo an extra round of genome duplication [6]. Comparatively, fewer unduplicated genes or singletons were found in these vertebrate species, with median values of 21% in mammals, 25% in birds, 22% in reptiles, and 18% in fish (Fig. 1B,D). No significant difference in the frequency of duplicates/singletons was found in any clade (Fig. 1C). There were a few notable exceptions. For instance, in fish, the fewest number of duplicates were found in the lamprey, which is understandable as the lamprey is one of the earliest vertebrates and diverged from other fish species before the third WGD event round [6, 25, 26]

Fig. 1.

Duplicates and singletons in vertebrates. (A,B) The frequency of duplicates (A) and singletons (B) in different vertebrate species is represented by boxplots. The bottom and top of the box illustrate the 25th (Q1) and 75th (Q3) percentiles, respectively, while the midline in the box marks the median value. The bottom and top of the whiskers refer to the 10th and 90th percentiles, respectively. Data points beyond the whiskers are considered outliers and are plotted individually as small circles. Outliers represent values that are significantly different from the rest of the data. The scale used for the y-axis is linear, representing values in percentages. The detailed data can be seen in Supplementary File 2. (C) Significance level for difference in duplicated genes. A z-test was applied to estimate the p-values at a significance level of 0.05. (D) The frequency of singletons in the representative species of each of the four vertebrate taxa is shown in a simple cladogram.

The presence of singletons in a considerable portion of human genes hints towards their evolutionary and physiological importance. The question arises: Are these singletons the evolutionary relics, represented as the sole survivors of an ancestral gene family that has undergone gradual gene losses and eventually ended up as a single gene? Or are singletons, or most of them, representative of convergent evolution, resulting in the formation of new genetic toolkits in the form of novel genes in vertebrate genomes? An evolutionarily surviving singleton will most probably have orthologs. Conversely, it would be difficult for a de novo singleton to present in multiple species as the chances of forming protein-coding genes with similar genetic makeups in various species must be extremely low. In this regard, we attempted to determine the status of orthologs of human singletons (OHSs) in each vertebrate species in our dataset. We found that OHSs are present in numerous species, indicating that these result from divergent evolution and not merely novel innovations. Interestingly, a predominant majority of human singletons seemed lost in other species, with median values for percentage loss of 29%, 45%, 43%, and 48% in mammals, birds, reptiles, and fish species, respectively (Fig. 2A,C and Supplementary File 3). Thus, a substantial fraction of OHSs seems to be lost throughout the vertebrate clade. The significantly low percentage loss value for mammals (with a z-score of –2.36 and p-value of 0.01 at a 5% significance level) is perhaps due to less evolutionary divergence from mammals than other clades in our dataset. Likewise, because fish are most distantly related to mammals, the loss of OHSs is relatively higher compared to different clades. Notably, the median values for loss of orthologs of duplicates were relatively lower than the loss we observed for the OHSs, i.e., 21%, 39%, 38%, and 54% for mammals, birds, reptiles, and fish, respectively (Fig. 2B and Supplementary File 3). These findings contrast the common assertion that singletons evolve under relatively strict functional constraints compared to duplicates [27, 28]. Indeed, certain previous studies have questioned such assertions [9, 29].

Fig. 2.

Singletons, the final survivor in a gene family. (A,B) Loss of orthologs of human singletons and duplicates in vertebrates, respectively. (C) A simple cladogram shows the loss of singletons and duplicates of representative species for each of the four vertebrate taxa. The scores on the left and right sides of the slash represent the percentage loss of orthologs of human singletons and duplicates, respectively. (D,E) The frequency of orthologs of human singletons (OHSs) is for the singletons and duplicates, respectively, in different vertebrate species. For each boxplot, the bottom and top of the box illustrate the 25th (Q1) and 75th (Q3) percentiles, respectively, while the midline in the box marks the median value. The bottom and top of the whiskers refer to the 10th and 90th percentiles, respectively. Data points beyond the whiskers are considered outliers and are plotted individually as small circles. Outliers represent values that are significantly different from the rest of the data. The scale used for the y-axis is linear, representing values in percentages. The detailed data can be seen in Supplementary Files 3,4.

One way of looking at this loss of OHSs is not to see it as a loss of individual singletons but to consider them a part of the ancestral gene family. These singletons would have once been part of a gene family. The loss of their ancestral paralog genes means these gene families were already evolving under relaxed functional constraints. As a result, these ancestral paralogs were lost, leaving behind the singletons as the remnants of their ancestral gene family. Since the ancestral paralogs did not provide a survival advantage, it is intuitive to think that these singletons, or at least some of them, also evolved under relaxed functional constraints and, thus, did not provide a significant survival advantage. Therefore, it was only a matter of time before these singletons faced the same fate as their paralogs. However, the question arises of how a gene family undertakes such a drastic evolutionary route. One answer may be that many gene families are part of a supergene family; hence, they may share potentially redundant protein domains in that superfamily, ultimately promoting the loss of most of the members in that gene family [30]. This potential process also underscores the importance of more rigorous and comprehensive functional and evolutionary studies to evaluate the evolution of singletons. These genes are not unique but also have the potential to provide us with a window into the evolution of the ebbs and flows of gene families. Interestingly, the dominant frequency of the OHSs in different species remained singletons, with a median value of 99.2%, 99.8%, 99.4%, and 98.7% in mammals, birds, reptiles, and fish species, respectively. Consequently, a very negligible frequency of genes, with a median value of 0.8%, 0.2%, 0.6%, and 1.3% in mammals, birds, reptiles, and fish species, respectively, were non-singletons or had paralogs (Fig. 2D,E, and Supplementary File 4). These data show that numerous OHSs are lost (as shown in Fig. 2A), and the tendency of vertebrates to lose singletons is higher than their expansion capacity. Therefore, gene loss is more prevalent than gene gain in vertebrates, irrespective of whether genes are duplicated or singletons. The loss of duplicated genes leads to the compaction in gene family size; sometimes, this compaction is so extensive that all the members of a gene family are lost except one surviving member, known as the singleton.

To see to what extent OHSs are present in different species, we determined the presence or absence of each OHS in individual vertebrate species in our dataset. Our findings demonstrated that not only do the majority of the human singletons have orthologs in numerous species, but the largest majority of orthologs were present in the predominant majority of species in our dataset, with the largest majority of OHSs ($>$43%) present in almost all of the species (Fig. 3A and Supplementary File 5). Intriguingly, our data especially highlighted those OHSs present in as few as only one species, i.e., in addition to humans, these genes are present in only one other species. This finding led us to believe species-specific singletons must also be present only in that species. We labeled this special set of singletons present in only one species, ‘super-singletons’. These super-singletons need some special attention because they are the sole representative of the genetic makeup they contain. Other studies have also mentioned the super-singletons as orphan genes [31, 32]. However, naming them orphan genes usually implies that these are novel genes that might have arisen through de novo synthesis [31, 33]. Thus, we used the term super-singletons here to underscore those singletons that are evolutionary leftovers, which solely represent a distinct genetic makeup in the biosphere. These super-singletons represent an evolutionary history of the substantial loss of gene members in the ancestral gene families but also represent those genetic relics that can be extinct from the biosphere. For instance, what about the species on the verge of extinction? Following the extinction of these species, the genes and the proteins encoded by them would not be lost, as orthologs of these lost genes would thrive in other species. However, this would not be the case for super-singletons. An extinction of any species that harbors these super-singletons will also lead to the extinction of a whole set of genes. Here, we have shown the super-singletons present in 10 endangered species. These super-singletons present in these 10 endangered species are the most probable candidates for future extinction of genes (Fig. 3B and Supplementary File 6). Previous studies have focused on the genetic makeup of species that have become extinct or are now endangered [34, 35, 36]; however, this study presents an alternative scenario whereby the extinction of species would lead to the extinction of a set of genes: the extinction of super-singletons. Thus, this study will document a species extinction event as the disappearance of a species from the biosphere and the extinction of a gene or a set of genes. Gene extinction would be a different event than gene loss because, in the case of gene loss, genes are lost from one or two species, whereas gene extinction leads to the disappearance of genes from the biosphere. In the case of gene loss, the orthologous genes in other species remain available for all sorts of genetic manifestations; indeed, the lost gene in a given species may even be restored due to horizontal gene transfer and hybridization, etc. [37, 38]. However, for gene extinction, the gene is permanently lost from the biosphere, remaining only in previous literature. In addition, identifying super-singletons provides us a platform to understand the evolutionary forces acting upon gene families. There are certain gene families in which the functional constraint of the genes has relaxed to the extent that all the other family members have been lost in all the species except one. This has caused us to posit, what does this extreme level of attrition of gene family mean? Can these super-singletons also be a marker for other gene families or provide a hint about their fate? Are super-singletons generally present in endangered species, especially just a random event, or does it relate to species conservation? Is there a species-specific advantage that has allowed these genes to remain in a species, although they have been lost in every other species? Are there certain protein structures that these genes encode and which are on the verge of extinction? Finally, on a broader level, is there a link between the tendency of species genomes to lose genes and the species becoming endangered or even extinct? These are the questions that this study has made most pertinent, and their answers need to be explored to understand the evolution of gene families and the important but intricate relationship between genome stability and species survival.

Fig. 3.

Super-singletons and gene extinction. (A) The frequency of OHSs is present in a given number of species. Values on the x-axis represent several species, while values on the y-axis represent the number of OHSs present. Please see Supplementary File 5 for details. (B) Super-singletons in endangered species. The figure shows the Gene IDs of 10 super-singletons in five representative endangered species. The images of the species are taken from Wiki Commons (Chimp, T.devil, P.turtle, B.baby, Tagua). For a complete list of super-singletons in all the endangered species in our dataset, please refer to Supplementary File 6. Abbreviations: T. devil, Tasmanian devil; B. baby, bush baby; P. turtle, painted turtle.