Thanks to Babbel, a language learning app, for supporting this SciShow video.

As a SciShow viewer, you can use our link to grow your language skills with Babbel for up to 60% off with a 20 day money-back guarantee. Solving genetic diseases is hard.

Like, there are a ton of genes and gene variants, and picking through them all one by one to find out which ones are troublemakers is like looking for a needle in a haystack. It’s such a problem that geneticists have been kind of stuck for awhile - after the completion of the Human Genome Project, there’s only been so much progress made with the data available. Researchers figured they may be able to pick through that hay a lot faster if they widen the search net a little.

So some geneticists interested in curing human diseases have been exploring the DNA not of humans, but of a huge group of animals. Specifically, our fellow primates. [intro song] This project has been called the Primate Genome Project, for obvious reasons. It involved more than 100 researchers based in countries all over the world.

One of the groups involved in all this cataloged DNA samples from more than 200 kinds of primates, representing around half of the primate species out there. Some blood or skin samples came from animals in captivity, while others were taken from animals living au naturel, in the wild. And some of the DNA came from dead specimens from zoos and other institutions that were collected as far back as the 1960s.

Getting ahold of tissues from all of these species was a huge undertaking, largely supported by Illumina, a DNA-sequencing company that usually deals with genetic material of the human variety. The story goes that they initially proposed the project with the idea that the company would cover the sequencing in return for the hard-won samples, and the resulting data would be shared all around. But this wasn’t just fueled by a passion for primate genetics research.

See, despite the revolution in DNA sequencing technology, the field of human genetics research has been an uphill battle for quite a while. So scientists decided to cast a bigger net to capture genome data from nearly half of all primate species with the hopes of gaining new insights into age-old problems. The Human Genome Project, which began in 1990 and led to the publication of the first nearly-complete human genome sequence in 2003, came with, like, some high hopes.

Researchers had dreams of deciphering the entire make-up of our genetic code and finding lots of new approaches to predict, prevent, and treat disease without harmful side effects. But even though we know that genetic mutations cause diseases, actually finding the specific genetic mutations to blame has been challenging. Because our genomes are so immensely long and complicated, it’s pretty much inevitable that a few errors get made every time the DNA is replicated.

Which adds up when you consider that a human’s genome is about three billion characters long. . Part of what makes any individual human unique is these differences , what we know as mutations, or variants. Most genes instruct our cells to build molecules called proteins that do all kinds of things in our bodies, like helping us digest food and creating pigment in our skin or hair… That’s just to name just a couple of things. .

But sometimes when genetic mutations mess up the code, a really important protein with a crucial function might get built incorrectly, or might not get built at all. When these mutations occur in the cells that become sperm and egg, they’re heritable, meaning they can be passed to the next generation. Which is how we end up with genetic diseases like, say, Huntington’s Disease.

Huntington’s is actually a really simplified example, since it’s controlled by mutations in just one gene. Most human diseases are complex and stem from mutations in lots of genes, plus how those genes are expressed, and environmental factors, too. Which is why finding out which mutations are contributing or not contributing to disease has not been easy.

There are over 70 million predicted possible protein-altering variants in the human genome. But so far, only about 0.1% of those variants have known links to diseases. The rest are considered variants of “uncertain clinical significance” - meaning, we’ve got no idea if they’re harmful or not.

And This new idea was to get rid of a little of that uncertainty by comparing these mutations to a much larger dataset. So, now that we have all these primate genomes , what are we gonna do with them? Essentially, we need to play a massive game of “spot the difference.” The researchers needed to identify genes that are stable, or almost never vary across all of these species.

The idea is that if a gene almost never varies, that means mutations in it would be like, really bad. And the opposite is also true: If a specific mutation shows up in lots of other species, you can infer that mutation probably doesn’t mess with any important functions even if it tweaks a protein. Basically, not that bad.

So when they play "spot the difference" and find a gene variant in multiple species, that gene gets crossed off the list of suspects for being harmful. Using this model, they flagged over 4 million variants that seemed likely to be benign because they showed up in lots of other primates, too. But there’s still a lot left to do.

Like we said, there are more than 70 million variants to pick through. And this is where AI comes in. The team used the primate variants to train a neural network that they named PrimateAI-3D.

It analyzes the structure of proteins to figure out whether mutations in the code mess it up enough to cause problems. These researchers used scores from PrimateAI-3D to identify genes that had a higher association with disease when compared to gene variant data of more than 450,000 humans. And the specific messed-up genes that the program identified allowed researchers to hone in on potential causes of complex health issues..

One example is high cholesterol. The team identified certain rare variants in two genes, LDLR and PCSK9, that probably affect your odds of developing high cholesterol and heart disease. Previous work had identified PCSK9 as a gene of interest, but hadn’t figured out if this specific variant was a major player or not.

So having this evidence gives researchers more ammo to investigate both genes’ roles in heart disease. They believe it could even be used to make targeted therapies for people with the mutations. And even though the main goal of the Primate Genome Project was to help solve human diseases, the unprecedented breadth of data gathered by both this team and others, has led to far-reaching insights about other primates, too.

The newly revealed genomic relationships between these hundreds of species tell stories of primate migration and adaptations. Scientists are hopeful that it can be used in conservation efforts, too. For instance, analyzing these primate genomes has contributed to the body of science that challenges the somewhat dated idea of immutable species.

When we think about a species, the most simplified definition is that two animals are in the same species if they can reproduce and make fertile offspring with each other. The classic example is that even though a horse and a donkey can breed to make a mule, they’re not the same species because the mule is almost always sterile. Research has increasingly shown that the whole idea of a species doesn’t always work that way.

Hybrids, as in the offspring of two individuals that are in totally different species, can occasionally be fertile. And this new genome data revealed that primates can and often do have fertile hybrids, too - enough to see the evidence of them in their genomes. That tells us a ton about what these little guys get up to in the wild, where their territories overlap, and also… like, how they see their like, species cousins, so to speak.

The research is also revealing details about the effects of varied habitats on genomic change in species. This knowledge could help with conservation efforts of threatened primates. We’re continuing to learn about ourselves by investigating other primates with these new approaches, uncovering more than we knew before.

There is still a lot more to learn from these primate genomes, both about us and about them. But thanks to our evolutionary cousins, we’re picking through that haystack faster than before! Thank you to Babbel for supporting this SciShow video!

Babbel is the #1 language-learning app in the world, and it got that title by keeping things flexible to work with your schedule. If all you have is 10 minutes in a day, you can use Babbel to learn an entirely new language in increments as short as 10 minutes. Then, after three weeks, you could start speaking Norwegian or Portuguese or Indonesian.

And after two months, you could be speaking all three! Imagine walking into any room in the world and being able to understand what people are saying. Babbel sets you on your way to actually having that super power.

And as a SciShow viewer, you can get up to 60% off when you sign up using the link in the description down below. Choose from different subscriptions, including a lifetime option, that comes with two free live classes. Thank you for learning with Babbel and with SciShow! [ OUTRO ]