We sequenced the whole genome of a phenotypically unique gorilla, identified and characterized the causative mutation for his albinism, and explored the origin of this trait. We found a private non-synonymous substitution in one of the candidate genes - the SLC45A2 gene - associated with the OCA4 class of albinism. We provided several lines of evidence based on evolution, human disease and a functional assay supporting that this mutation in a transmembrane domain can modify the topology of the translated protein, therefore reinforcing its causative role in this rare case of albinism. Moreover, long runs of homozygosity in this wild born individual explain the emergence of this recessive trait through identity by descent, suggesting that inbreeding was an important factor towards the emergence of this phenotype.

We inferred that Snowflake was an offspring of closely related individuals supported by an inbreeding coefficient of 0.118. In general, inbreeding is avoided in the wild because the offsprings within gorilla societies disperse to other groups before maturity [28]. This is strictly true in patriarchal groups that are commonly composed of a silverback male and several females (97% of all the gorilla groups) [6] whereas in multimale groups, females can remain and have their first birth in the natal group. Multimale groups are usually composed of related males and therefore, newborn females are likely to be also related to them (commonly with relationships such as half brothers or half uncles). However, multimale groups have mainly been observed in mountain gorillas, while only two multimale groups have ever been reported in Western lowland gorillas, suggesting that they are extremely rare in these populations [6]. Therefore, it seems unlikely that a multimale group would explain the inbreeding found in this Western lowland gorilla.

Previous parentage studies in wild Western lowland gorillas have never found inbred mating, suggesting that is probably a rare behavior [29]. Despite this and considering that the observation of inbreeding in a single individual could be an extreme case, some social observations may point that inbreeding may still occur. First, gorillas seem to follow a patrilocal social structure, i.e. silverbacks are usually related to one or more nearby silverbacks [30]. Additionally, females transfer several times during their lifespan after the dispersal from their natal group [31], which may result in the arrival of a female to a new group where the silverback is related to her. Although father-daughter inbreeding is completely avoided, this hypothesis is feasible because other mating relationships, with half-brothers or even full brothers, are possible; suggesting that females do not detect consanguinity [6]. Other factors such as habitat loss, small population sizes and population fragmentation may influence the disposal of breeding groups and therefore of unrelated silverbacks which may in turn favor inbreeding [32]. Other potential explanations are less likely; male takeovers are highly avoided and the death of a male silverback normally results in the disintegration of the group and female dispersal.

A previous study using microsatellite markers in captive gorilla populations showed that their genetic diversity is comparable to wild gorilla populations [11]. However, in our study, Snowflake shows different patterns of heterozygosity compared to the captive born gorillas.. The gorilla studbooks show that Kamilah (Studbook ID: 661) is a first generation captive-born gorilla, while Kwan (Studbook ID: 1107) is a second generation captive-born gorilla. When we compared the heterozygosity genome-wide, we observed that Kwan is the gorilla with higher heterozygosity, despite we cannot rule out that this was a result of some false positives due to the lower sequencing coverage. Kamilah and Snowflake have lower heterozygosity, with the albino gorilla showing the lowest values compared to the other captive-born individuals (Figure 3b) even accounting for the regions of inbreeding. This suggests that breeding programs could result in an increase of genetic variation but a bigger sample size would be needed to systematically explore this effect.

In this particular study, we show that high throughput sequencing can be used not only to unravel the genetic mechanisms of fundamental phenotypes (including disease) in non-model organisms, but also to provide insights into conservation genetics through the detection of inbreeding of endangered species such as gorilla. However, in order to systematically explore relationships and breeding patterns from wild specimens using whole genome sequencing data, high quality DNA is required and in most field studies, non-invasive samples such as feces or hair are used, and the amount of DNA extracted from such samples precludes the application of this methodology to conservation studies. Still, it can be applied to analyze the genomes of wild born individuals in zoos where blood samples are usually taken during routine veterinary check-ups. However, sequencing technologies quickly and constantly improve, and recent developments that includes library construction with very little amounts of DNA [33] or single-cell sequencing [34, 35] may allow the implementation of this kind of analyses into conservation genetics in the near future.