Has a company really discovered a million new species?

This BBC podcast is supported by ads outside the UK. Thank you for downloading the podcast. More or Less podcast. With a programme that looks at the numbers in the news, in life and in DNA databases. I'm Tom Coles. How many new species are out there waiting to be discovered? Quite a few, according to a claim that loyal listener Vivian heard in a BBC interview and wanted us to look into. Here's Dr Oliver Vince, co-founder of a company called Basecamp Research, speaking to the Today programme. Far more than 99% of life on Earth is completely unknown. And so Basecamp set out to change this. We've built partnerships in 28 countries, 150 different locations, as is the rainforest, the Arctic, volcanoes, under the ocean, to basically look for the rest of the species. So far, we've discovered over a million new species. A million new species? That does sound exciting. It also sounds like an awfully big number. Can it really be right? So my name's Rob Finn. I work at the European Biothermatics Institute. We are the custodians of the world's biological data. So we try and gather it all up and put it into databases and then make it freely available for the scientific community to use in their research. Although they do things in a different way and not for profit, Rob's databases are similar to the one being created by Basecamp Research. So he knows what this claim is all about. First things first are we talking here about the discovery of a colourful new species of frog or bird or a mysterious new plant No it the short answer Most of the times we talk about species you thinking about a large organism that you can see like a plant or an animal. However, what they're talking about are microbes, so things that you can't see with the naked eye, you need a microscope to see. These million new species are bacteria, microscopic single-cell organisms which multiply by dividing themselves in half. The bacteria you might have heard of, for all of the wrong reasons, are E. coli and Salmonella, which cause upset stomachs. But that is just the tip of a very big iceberg of very small things. There are a lot of bacteria out there. It's estimated there's about a trillion different species, which is the same number of stars as there are in the Milky Way. So there's a lot out there to be discovered. So the species bit is not what you might imagine if someone told you they'd discovered a new species. But the discovery bit is also not what you might imagine. You might think it involves thousands of scientists growing colonies of interesting microbes in little dishes and then watching them multiply through microscopes. but that's not how these species were discovered. The method scientists use is based around that special stuff that contains the blueprint for the organism. It's DNA. Every microbe has its own genome, a long sequence of DNA subdivided into small sections called genes and you can use the genome to identify the organism. With DNA sequencing getting cheaper and cheaper and computational approaches getting better, What we can now do is take environmental samples where you might have hundreds of different bacteria growing and we can sequence those all together. Basically, you scoop up water, mud or ice at the location you're interested in and then process the DNA you find there in bulk. So if you want to go and study the bacteria living in a marine environment, what you can do is you can take, for example, five litres of water and then you filter that through a really, really fine filter. So the water passes through, but all the microbes get caught up on this filter paper. Then and please don try this with other species You whiz it up You literally whiz it up You can either use sonication but that tends to break down the DNA or you can use a very gentle beating system that breaks open the cells So you are essentially putting it in a blitzer and whizzing it up, but in a gentle way. You pop open all of those single cell organisms and make a kind of DNA soup. And then we can pass those through a DNA sequencer that reads off their genetic code. You turn the physical soup into a data soup, in which those long strands of DNA are broken up into tiny fragments, all mixed in together. It's like giant jigsaw puzzles, and we have many different jigsaw puzzles that we must piece together again, and we use computational approaches just because of the size of the data sets and the complexity of the problem. I'm going to oversimplify an incredibly complicated process here, but there are rules and patterns in DNA, which means you can take that jumble of puzzle pieces and use computing power to start to identify the pieces that fit together. You end up digitally reconstructing the genomes, at least in part, of the original organisms that you started with. okay you've collected the samples made a dna data soup and identified the genomes of different organisms final part of the puzzle how do you know that these genomes represent different species so in animals and plants there's quite a nice definition about how you want to sort of talk about an individual species. So we understand species in the fact that they can interbreed in nature and produce fertile offspring. A horse and a donkey, for example, are different species but can still mate. They produce a mule as offspring, but mules cannot have babies themselves. Also, horses and donkeys look and sound very different. the problem is is that first of all bacteria do not breed in the same way that that these large organisms and there are so many of them that it becomes very difficult to look at every single one of them and come up with some characteristic that says oh this is this particular organism and this is another organism. Instead of using breeding and observable characteristics biologists use statistics to calculate how many species they found If we have two different genomes and then we can compare them so we can basically line up the two strands of DNA and then we can say how similar are they So if they are greater than 95% identical, that's the rule of thumb, then we would consider them to be the same species. now depending on what exact percentage you use in this analysis you get a slightly different number of species but overall rob says that while it's hard to check precisely the one million species calculation from base camp research looks in the right ballpark i think it's perfectly feasible from what they've done the environments they've gone into and how they've done it it is exactly that. Right ballpark, I might have done it differently, but it's there or thereabouts. The strange thing about this process, where you turn living organisms into data, is that the species only exist in quite an abstract sense. It might be useful for training an AI system on genetic code and potentially finding new and powerful medicines, but no one's actually seeing these species of bacteria alive or growing them in a lab. The approach of taking everything and then whizzing it up actually destroys the sample. So they're technically finding new species only by destroying them which seems how can I put it suboptimal but Rob thinks it's all worth it. I think there'll be some interesting discoveries no doubt and it's a great data set for them to exploit and for us to investigate collectively as scientists. Thanks to Rob Finn. That's it for this week. If you've seen the number you think we should take a look at, email us on moreorless at bbc.co.uk. Until next week, goodbye. Bye. Listen on BBC.com or wherever you get your podcasts.