Recombination-aware, whole-genome analyses of sarbecoviruses show that genomic fragments very closely related to SARS-CoV and SARS-CoV-2 circulated in horseshoe bats only years before human emergence. Phylogeography places recent ancestors in western China and northern Laos and indicates movement patterns inconsistent with bat-only dispersal, implicating intermediate hosts or wildlife trade.
0:00Welcome to Base by Base, the papercast that brings genomics to you wherever you are. Thanks for listening, and don't forget to follow and rate us in your podcast app. Glad to be here for another deep dive.
0:10So to start off today, I really want you to imagine a vast, just incredibly rugged landscape. We are talking about 1000s of kilometers of dense forests. Yeah, jagged mountain ranges, winding rivers. Exactly.
0:25And these ancient limestone caves just stretching across the rural expanse of Southeast Asia. It is an incredibly remote, you know, deeply complex ecological environment. Right. And deep within those cave systems, virtually isolated from human infrastructure for millennia.
0:39Microscopic entities are circulating in the dark. Just quietly passing between colonies of bats. It really paints this picture of something ancient and largely static. It does. Yeah. But then, we hit a profound mystery.
0:52What really happens when a microscopic entity, something that has been isolated in a remote limestone cave for centuries, suddenly appears in the middle of a bustling concrete magacity. Well, it creates a massive logistical paradox.
1:05Yeah, and I really want you to think about the logistics here. Because we usually look at outbreaks and ask, you know, how do viruses mutate to infect humans? Right. The biological side. But instead, let's think about the actual physical transit.
1:19How does something without legs cross a continent faster than its natural host can even fly? That is the big question. How does a global event spark so incredibly fast when the biology of the animal carrying it tells us that a journey of that distance should take centuries?
1:34That is the ultimate question of nature versus human infrastructure. It really requires us to map out the exact pathways these pathogens take from a remote cave to an urban center. Mapping both time and physical space.
1:47Exactly. And cracking that puzzle is the mission of this deep dive. Today we celebrate the work of Jonathan E.P.Car, Spiros Litris, Mahangafari, Michael Warud, Jolo Worthmeim, Philip Lemay, and their vast network of international collaborators, who have advanced our understanding of the recency and geographical origins of a bat virus's ancestral to SARS Cove in SARSCOVI2.
2:09It's a truly massive collaboration. It really is. And this open access research was published in the journal cell on June 12, 2025. They tackled a problem that has haunted molecular epidemiology for a very long time.
2:23Oh, absolutely. For decades, we've had a general idea of the starting point, but the middle of the map was effectively blank. Okay, let's unpack this. Why are we doing this deep dive and why was filling in that map such a massive problem?
2:36Well, we have known for a while that horseshoe bats are the natural reservoir for cervicoviruses. Right, and that's the specific subgenus that includes both SARScoV1, which triggered the outbreak back in 2002, and SARScoV2, which emerged in late 2019.
2:51Exactly. So we have the natural host identified. We know the bats carry these viruses. But tracking down the exact genetic ancestors of the specific viruses that spilled over into humans has been historically incredibly difficult.
3:04Wait, I mean, if we know they come from bats. Why couldn't researchers just sequence a bunch of bad viruses put them next to the human virus and just trace the family tree back to a common ancestor? It sounds easy enough.
3:15Right. That seems like the straightforward, logical 1st step. It seems logical until you account for how coronaviruses actually replicate. They do not just mutate by making simple copying errors. Oh, he's doing something else.
3:27Yeah, they frequently recombine. Recombine. Okay. When 2 different lineages of a coronavirus infect the exact same cell in a bat, the enzyme that copies the viral RNA can literally just fall off one genetic track and reattach to the other.
3:41So they are actively swapping chunks of genetic material while they replicate. Continually. I mean they are highly prone to the swapping. So any single bat virus genome you isolate today is not a single linear family tree.
3:56It completely mixed up. Right. It is effectively a mosaic. A patchwork of different genetic histories pulled from different viral lineages over 1000s of years. If you try to run a comparison on the whole genome all at once, the data gets hopelessly blurred.
4:10You're looking at a dozen different evolutionary histories matched into one. You know, I like to think of the viral genome like a car built entirely from junkyard parts. Oh, that's good way to look at it.
4:19Yeah, so if you just look at the whole car sitting in the lot, you cannot date it accurately. Because the engine might have been pulled from a 1995 sedan, the doors are from a 2014 hatchback and maybe the steering wheel is from a truck from the 1980s.
4:34Right. So if you try to average the age of the engine, the doors, and the steering wheel to find out when the car was built, you get a completely meaningless number. Exactly. You might conclude the cars from 2003, but no single part of the car was actually built in 2003.
4:49You just average the noise. Which is exactly what happened in previous studies. Earlier attempts to date the common ancestor of SARScoV2 and its related bat viruses, tried to look at long genomic fragments.
5:01So they were looking at the whole car. Yes. Because they were unknowingly averaging the ages of different recombined parts. They ended up estimating that the human virus' ancestor existed maybe 30 to 40 years ago.
5:13Wow. So to find the real origin, you have to computationally pull the car apart. You have to separate the engine from the doors. And that's the core methodology here. The researchers in this study moved entirely away from looking at the whole genome.
5:26What did they do instead? They focus their efforts on finding non-recombinant regions. They call these NRRs. In RRs, got it. They used highly sophisticated computational models to scan the viral genomes and identify the specific breakpoints.
5:42You mean the exact spots where the genetic sequence had been swapped? Exactly. By finding the seams, they could slice the genomes into clean segments that have not recombined. Meaning each of these individual NRRs shares a single uninterrupted evolutionary history.
5:57Yes. They basically treated the virus as a collection of independent genetic artifacts. For the virus's closely related to SARS CoV1, they identified 31 distinct non-recominant regions. 31 parts. And for the SARS Covi 2 like viruses, they found 44 NRRs.
6:14So breaking the genome down into 31 and 44 distinct pieces allows them to analyze the true age of each piece completely on its own. Exactly. But how do they actually determine the age of these individual fragments once they have them isolated?
6:27Well, for each NRR, they utilized a specialized molecular clock. A molecular clock. Yeah, in biology, a molecular clock relies on the mutation rate of biomolecules. Think of it like a metronome that ticks at a relatively steady pace.
6:41By counting the number of genetic differences, you know, the mutations between a bat virus fragment and a human virus fragment, you can calculate backward. To figure out the exact time in prehistory, when those 2 sequences diverge from a common ancestor.
6:57Right. But counting mutations over 1000s of years seems risky. What happens if a mutation happens twice in the exact same spot? That is exactly the issue. I mean, I'm thinking of substitution saturation.
7:10It sounds like when you write a math problem on a whiteboard, and then someone writes over it and writes over again until it's just a smudgy mess. Substitution saturation is the absolute bane of deep time evolutionary biology.
7:21Over a few decades, a virus might mutate at a specific spot, changing an A to a G. We can read that easily. But if you stretch the timeline to 100s or 1000s of years, that same spot might mutate back from a G to an A or flip to a C.
7:37Ah, so the new mutations overwrite the older mutations entirely. Completely. So if you are simply counting the visible differences between 2 sequences to calculate how much time has passed, you are vastly undercounting.
7:49Because you are missing all the invisible overwritten changes. The mentronome looks like it's ticking slower than it actually is. It throws deep time estimates completely off. The researchers news, standard molecular clocks would fail here.
8:03So to correct for this massive problem, they applied something called the prisoner of war model, or poud W for short. Okay, I am stuck on the name. How does a prisoner of war model solve the smudgy whiteboard problem?
8:15The PowW model is a highly advanced mathematical framework that specifically accounts for time dependent evolutionary rates. Meaning what, exactly? It deals with the biological reality that over short periods, you see a lot of mutations, many of which are actually harmful to the virus, but over long periods, natural selection purges those harmful mutations from the population.
8:37So the mutation rate looks really fast in the short term, but artificially slow in the long term, because the bad mutations just disappear from the historical record. Exactly. The Pow W model essentially acts like a forensic accountant.
8:52It doesn't just look at the final tally of mutations. It looks deeper. Right. It corrects the curve of the clock based on how long natural selection has had to act on the sequence. It mathematically compensates for the invisible, overwritten mutations.
9:06Which allows the researchers to trace the lineage of these viral fragments back 1000s of years without losing accuracy. Precisely. Okay, so once they scrubbed away those evolutionary smudges using the PWW model.
9:18They finally had a clean timeline for every single piece of the virus. They did. But time is only half the logistical puzzle. We still need to know where these viral pieces were actually circulating. What's fascinating here is how the team essentially built a multi-layered historical map.
9:33They mapped the evolutionary tree of each of those 31 and 44 fragments, and then they layered on phylogeography. Phylo Geography being the intersection of genetics and physical geography. Yes. They took the physical GPS coordinates of where all these related bat viruses have been sampled across Asia over the years.
9:51Okay. And they built a spatially explicit model to see how fast the different viral lineages physically moved across the landscape over time. So they're tracking the trade routes of the engine, the doors, and the steering wheel of our junkyard car, all simultaneously.
10:06That's exactly it. And when they finally ran this massive complex model across all those distinct NRRs, the timeline completely collapsed. Wow. The results here completely shift the narrative we thought we knew.
10:17Let's start with the recency, because remember how those earlier flawed estimates, the ones that average the whole genome, suggested the ancestors of these human viruses circulated decades ago. Right, 30 or 40 years.
10:29But when this team looked closely at the clean NRRs using the corrected molecular clock, they found fragments that were incredibly recent. Shockingly recent, the closest ancestors of the viruses that spilled over into humans existed less than a decade before the outbreaks.
10:44Less than a decade. Give me the exact dates they uncovered. For SARSCOVI one, which caused the global outbreak starting in 2002, the model found that one of the non-recombinant regions traced back to a bat virus ancestor circulating in 2001.
10:59Just one single year prior to the human outbreak. That is wild. What about SARS CoV2? Well, SARS CoVI 2 emerged in human populations in late 2019. The POW model identified 2 distinct NRRs whose closest ancestors circulated in bats in 2014.
11:15That is merely 5 years prior to the pandemic. Yes, 5 years. That completely shatters the prevailing theory that these viruses were slowly brewing in some unknown intermediary animal for 30 or 40 years.
11:26It really does. The ancestral pieces of these viruses were sitting in bat caves, just a handful of years before the outbreak started. The recency is startling. But when you overlay the geography, the logistical paradox really emerges.
11:39Yeah, through their phylogeographic mapping, they pinpointed the physical locations where these recent ancestors were circulating. Where were they For SARSCOVI one, the closest ancestors were circulating in Western China, specifically in rural, mountainous regions like Yunnan, Sichuan, or Gizu.
11:56But Tsars Kovi 1 didn't emerge in Western China. It emerged in humans in Guangzhou in Guangdong Province. Which is roughly a 1000 kilometers away from those caves. And the pattern repeats almost identically for SARS Kovi too.
12:09Yes, the phylogeography shows its closest ancestors, circulated in Yunan, China, or Northern Laos. Yet the human outbreak ignited in Wuhan, in Hubei Province. Once again, over a 1000 kilometers away from the natural bat reservoir.
12:23Exactly. Here's where it gets really interesting. You're telling me these viruses crossed over a 1000 kilometers of rugged terrain in just a few short years to reach Guangzhou and Wuhan. That is the crux of the issue.
12:34How fast is this virus naturally travel on its own? Well, the phylogeographic analysis revealed a very strong statistical signal of something called isolation by distance. What does that mean in practice?
12:46This essentially means the viruses naturally travel across the landscape at a speed that directly matches the physical movement of their hosts. So the virus can only move as fast as a horseshoe bat can fly.
12:57Precisely. And I am guessing horseshoe bats are not marathon flyers making cross country tracks. Not at all. They are notoriously local. Horseshoe bats have a very limited dispersal capacity. They are not migratory birds.
13:11They stick to their local karst formations in cave systems. foraging within just a few kilometers of their roost. So the virus spreads incredibly slowly in nature, jumping from one bat colony to a neighboring bat colony over the course of generations.
13:26Exactly. It takes a massive amount of time for a lineage to move 100s of kilometers across the map. So we have a massive mathematical contradiction here. The geographic data shows the virus was deep in the caves of Yunnin or Laos.
13:39The genetic dating shows it was there just 5 years before it showed up in a major city, a 1000 kilometers away. Yeah. But the biology of the bat proves it would take vastly longer than 5 years to naturally carry the virus that distance.
13:53It is a biological and mathematical impossibility for natural bat dispersal to account for that movement. The bats simply do not fly fast enough or far enough to bridge that geographical gap in that incredibly short timeframe.
14:07Which leads to the only logical conclusion, the virus had to hitch a riot on something else. If we connect this to the bigger picture, this finding strongly underscores the necessary role of the intermediate wildlife trade.
14:19We are talking about the massive human mediated transport of susceptible animals. Animals like palm civets and raccoon dogs, which we know can easily contract these viruses. Exactly. If these animals are farmed or caught in the rural regions where the bat reservoirs are located, they can be exposed to the virus.
14:35And then they are loaded onto trucks and transported across the country to densely populated urban markets in a matter of days. They act as a massive accelerant. They are basically the high speed rail for the virus.
14:47They bypass the natural bat speed limit entirely. Human driven networks, not natural bat migrations, are the true catalyst for these zoonotic spillovers. That's incredible. The wild and farmed animal trade provides the logistical bridge that transports a slow moving cave dwelling pathogen directly into the heart of a major human metropolis.
15:09So we are the ones bridging the gap between the remote cave and the city center. Unfortunately, yes. It is a chilling realization. The human infrastructure is the primary mechanism of emergence, but this raises an important question regarding the research itself.
15:22What's that? If these ancestral viral fragments were circulating so recently in 2001 and 2014, why can't researchers just go to a cave in Yonin or Laos right now, sample the bats and find the exact 100% genetic match?
15:37Yeah people ask that a lot. Why can't we find the direct ancestor alive today? Well, because evolution doesn't stop. It's a key limitation of the study and of genomics in general. Right, time keeps moving.
15:48We will likely never find the exact, intact, direct ancestor in a wild bet today. Since 2002 and 2019, Those viruses have continued to live, mutate, and critically recombine in the wild. Ah, so the viral landscape in those cases today is completely different than it was 5 or 10 years ago.
16:06Exactly. Going back to your analogy, the car has been dismantled again. And its parts have been put into entirely new vehicles. Wow. That is why future sampling efforts cannot just look at small fragments of the virus.
16:18In the past, researchers would often sequence just a small piece of a bat virus, like the RDRP gene, to see if it was related to SARS. But this study proves that you need to sequence the entire genome of every sample.
16:30Yes. Only by having the whole genome can you computationally break it into all its different NRRs. Only then can you capture the full mosaic history of the virus before the parts get scattered again. So why does this matter for you?
16:44Let's distill this down to the core insight. By treating viral genomes as a mosaic of different evolutionary histories. Researchers discovered that the bat ancestors of SARS Kovi 1 and SARS Kovi 2 circulated just years prior to their respective outbreaks, 1000s of kilometers away from where they emerged.
17:02Because the viruses naturally travel slowly alongside their bat hosts, human mediated wildlife trade was the necessary bridge that transported them to major cities. It forces us to look at the highway map, not just the microscope.
17:16The biology of the virus relies entirely on the logistics of human trade to cause a rapid emergence event. It really does. What does this mean for how we regulate and monitor the global wildlife trade in an increasingly interconnected world?
17:30Well, if human infrastructure is the primary accelerant, then surveillance cannot be limited to the destination, we have to monitor the transit routes. We have to secure the bridge, not just look at the cave in the city.
17:41Before we wrap up. I want to leave you with a final lingering thought based on these findings. We have established that these viruses have been circulating in relatively static bat populations for millennia, restricted by the slow movement of the bats across the landscape.
17:56Right. But what happens if global climate change suddenly and dramatically shifts these ancient habitats? It is a deeply unsettling prospect. As temperatures rise and extreme weather patterns alter the forests, the insect populations and resources those bats rely on will shift geographically.
18:14the bats will be forced to follow them. Exactly. If the bats themselves are forced to move out of their historical ranges and into new, highly populated agricultural or urban territories to survive, the natural speed limit of these viruses might change forever.
18:28We might not even need the wildlife trade to bridge the gap anymore. We might be bringing the caves right to our backyards. It is a stark reminder that we are not observers separate from this ecosystem.
18:38Our actions are actively re-engineering the map. This episode was based on an open access article under the CCBY 4.0 license. You can find a direct link to the paper and the license in our episode description.
18:50If you enjoyed this, follow or subscribe in your podcast app and leave a 5 star rating. If you'd like to support our work, use the donation link in the description. Now stay with us for an original track created, especially for this episode, and inspired by the article you've just heard about.
19:05Thanks for listening, and join us next time as we explore more science, base by base.
