 Andrew, you are not a part of this. American TV Daytime Talk shows are famous for two things, stripper fights and paternity tests. It's probably the lowest form of entertainment, but ironically it's a good illustration of how DNA technology has changed our society. The paternity tests, like this one from Identa Gene, are based on the presence of DNA markers. These markers are differences in the DNA sequence between any two individuals who are only distantly related. Because DNA is the basis of heredity, the number of shared markers is a good measure of relatedness. This is accepted by the scientific community, as well as by courts of law, and such profound authorities as Maury Povich and Jerry Springer. I want to talk about how these paternity tests work, how this relates to evolution, and I want to share with you the joys of phylogeny. So how do they work? There are a variety of ways to test for distinctive variations in DNA. The oldest are microsatellites, restriction fragment length polymorphism, or RFLP, and short tandem repeats, or STRs. These are still used, but they've been supplemented or replaced by didioxy DNA sequencing, amplified fragment length polymorphism, or AFLP, and a variety of other techniques, like SNP detection and real-time PCR. Polymerase chain reaction, or PCR, has changed this field dramatically. A commercial paternity test usually involves buckle swabs, which is a wooden stick rubbing inside of the mouth to obtain cheek cells, which contain DNA. The mother, the child, and the potential father should all provide the DNA samples. In general, the child's markers should be an overlay of half of the mother and half of the father's marker. There are some interesting exceptions, but not relevant here. The lab receives the samples and tests them for the presence of the markers. Then an algorithm calculates a probability of a match based on the shared markers of mother and father. The math is surprisingly straightforward. The probability is the product, that's multiplication by the way, of all the parent-age index values for each locus or individual site. The cost of a test is $149 from IdentaGene, and you can buy it at your local CVS for much less. The results come in three to five days and are 99.99% accurate. How close is that to the movie world of Gattaca? IdentaGene also offers tests for grandparents, aunts, uncles, and even for ancient ancestry. Ever wonder if you have East African ancestors? Now you can find out. Courts should make this mandatory for people convicted of racially motivated crimes, and they should film the criminal getting the results. I think we would see some very entertaining revelations. Jerry Springer could host. Back to the topic. We've seen that DNA markers or genetic changes can reveal relatedness. Biologists use the same process for classifying populations of organisms. Imagine you're going to the rainforest of Costa Rica. You're going to collect snails from the forest floor and attempt to create a classification system for them. Do you base it on morphology, shell size, color, location found, speed or skin texture? How relevant are those differences? It's a lot harder than it sounds. Enter DNA typing technology. Suddenly what looks like three distinct populations of snails turns out to be one genetically identical population of snail, but differing in nutrition based on local environment. The process is very similar to our talk show paternity tests, except that the snails don't call each other names and get into a slap bite on stage. So DNA markers are important for classifying species. This field is called phylogenetics and has largely replaced the old fashioned Linnaean taxonomy you learned in school. You memorized King Philip came over from Green Spain or whatever mnemonic you used for nothing. Don't blame me, blame your biology teacher. So where am I going with this? Well, I issued a challenge to the Discovery Institute to find a gene without homology or alternative function. It was picked up by Thunderfoot and P.C. Meyers. The response I got back from two different ID blogs was that homology does not imply relatedness. And in his response, Casey Luskin of the Discovery Institute referred to an article he wrote in which he dismissed phylogenetics and claims that the methodology for inferring common descent has broken down. Proponents of neo-Darwinian evolution are forced into reasoning that similarity implies common ancestry, except for when it doesn't. And when it doesn't, they appeal to all sorts of ad hoc rationalizations to save common ancestry. He then goes on to quote mine an article in The New Scientist, but omits the editor's note which states, none of this should give sucker to creationists whose blinkered universe is doubtless already buzzing with the news that new scientist has announced Darwin is wrong. Expect to find excerpts ripped out of context and presented as evidence that biologists are deserting the theory of evolution en masse. They are not. The editors of The New Scientist must be psychic. I propose to test the idea that phylogenetics lies in tatters and with it the basis for paternity testing. I'm going to show a series of phylogenetic trees I created using public domain data and online tools. In a separate video, I will show you how you too can construct phylogenetic trees without leaving your comfy chair. A brief background on phylograms. They reflect genetic distance graphically as a series of branches. I'll include a few links on the right on how to interpret them properly. Here's our first gene product, catalase. Catalase is an enzyme that protects cells from oxidative damage. We would expect it to be evolutionarily conserved, and it is. But we can see how organisms who are closely related are grouped together by the genetic markers in the catalase sequence. Here's what those similarities look like in what we've called a multiple alignment, a way of representing these genetic differences with one letter codes for each amino acid. The coloring code shows functional similarities between the different sequences. It's worth noting that if Casey Luskin is right, if the designer is reusing parts, she's creating an awful lot of different versions of the same enzyme to do the same job. That seems like a lot of unnecessary changes and diversity of forms for the same application. The designer, I suppose, must be inscrutable, so whatever we observe she must have meant for it to look that way. And that's fine for apologetics, but it's disastrous for a scientific theory. This is the short list of organisms that were present in homology. When I expand the scope of this to include a lot of other organisms which are not included in homology, I get a filogram that I can't fit on one page. I've had to break it down to be readable. I've got on here three hominids, three species of dolphins, six fish, three birds, two frogs, five plants, 12 insects, and a smattering of other groups just to make it interesting. When the animal name is followed by a number, that represents the duplication of the gene sequence. You can see that the duplications are more closely related to each other than they are to other organisms. So let's see how this breaks down by taxonomy. Well, the vertebrates are all grouped correctly, although I'm not sure how far back dolphins and mice share an ancestor. Let's break this down a little more. All the dolphins and the halmonids separately are all very close together, suggesting that they are genetically very similar and therefore closely related to each other. The two types of yeast are actually very distantly related, which I found surprising, but apparently this is well known to yeast biologists. It looks like the homology and taxonomy line up for all of these, which is really pretty neat. There is one organism out of place on this graph and it's a round worm that appears to be more related to bacteria than to other roundworms. What shall we make of this? I did a little research on it thinking I had discovered a flaw in the phylogeny. Actually, it's a flaw in my research. Photorabdus is a pathogenic bacteria of a roundworm, which is itself a pathogen of insects. I misread the description from the phylum. I wouldn't have counted if it hadn't been for the phylogran. So let's take a look at the second half of the phylogran for catalysis. I honestly don't know the taxonomy of some of these things, but I'll group the ones I know. Shrimps and prawns are closely related. The mosquitoes are all closely related. But the blueet or blueae, I don't know how you say it, dragonfly seems to be an outlier. I was very concerned that Casey was right, phylogenetics is in tatters until I did a little reading on dragonflies. They are very genetically diverse, often producing what are called cryptic species. And this particular one has more arthropod markers, making it appear more closely related to shrimp than a beetle for this one protein. I'm going to save others for a future video, but I hope I provide some information for proponents of intelligence designed to review. So does homology imply relatedness? I think the answer is obvious.