An intro to the neutral theory of evolution

From a top donor:

“I’d like you to write a blog entry (primer) about neutral theory aimed at the layperson.”

Okay, I’m not going to lie. I’ve been secretly hoping someone would bump this question out of the top ten, mainly because neutral theory is kind of boring and vaguely confusing and hard enough to explain while using biology buzzwords. It’s even harder to explain when you only have 30 minutes to write about it and it’s supposed to be targeting non-biologists. When I shared this question with some fellow genetics grad students, the general response was “Ewwwww.”

But I will try my best!

When most people think about evolution, they think of adaptations. Something in the environment puts selective pressure on a certain trait, and organisms with that trait are more “fit” (reproduce more). For example, rabbits that live in snowy climates are more likely to survive (and reproduce) if they have white fur that helps them blend into the snow. If a mutation randomly arises that make their fur white, or just lighter, that rabbit has an advantage over the other rabbits – the dark brown ones are going to be the first ones that are eaten.

There are lots of examples of adaptive evolution through natural selection, and people know them more because they make good stories. The most famous example of evolution, Darwin’s finches, is a case of adaptive evolution.

But not all evolution takes place because of natural selection. Evolution is at its simplest defined as the change in allele frequencies over time. That’s where neutral theory comes in. Neutral theory states that most evolutionary changes are the result of random drift of neutral mutants.

Buzzwords buzzwords, I know. So let’s take it one step at a time and pretend that we’re looking at a population of unicorns (if we’re pretending, might as well pretend all the way).

Alleles are just two different forms of a gene. Genes are usually hundreds or thousands of basepairs long, but let’s pretend we’re zooming in on three bases in a gene for fur color. If you have the sequence AAA, your fur is white. But if you have AAG, your fur is pink. AAA and AAG are different alleles.

Now let’s say all of the unicorns start as AAA, and then from mutation you get a unicorn who’s AAG. Being pink has no effect on the unicorn. He’s not more likely to get eaten, he doesn’t live longer, he doesn’t have more luck with lady unicorns. It’s a “neutral” mutation because it doesn’t change the unicorn’s fitness.

That’s where random drift comes in. Drift simply refers to the frequency of one allele changing due to random chance. That is, nothing is selecting for pinkness. Maybe that unicorn just happened to have more offspring. Maybe the population underwent a bottleneck and was reduced to just a few unicorns, the pink one happened to survive, and now his pinkness will make up a larger percentage of the population. Maybe a couple of unicorns, including the pink one, happened to get isolated on one side of a river, so that side eventually had a lot of pink unicorns, while the other has a lot of white ones. Through random chance alone, a neutral mutation can grow to high frequency or even reach 100% (what biologists call “becoming fixed in a population.”)

When you get into the mathematics, you assume that random neutral mutations occur at the same rates across individuals. This is how biologists get things like “molecular clocks” where they can tell when two species diverged from each other.

…And I have no idea if that made any sense, but I’m out of time. You have all now been exposed to (a shoddy summary of) evolutionary theory, congratulations. If anyone would like to explain further or correct me in the comments, please do so!
This is post 20 of 49 of Blogathon. Pledge a donation to the Secular Student Alliance here.

Neanderthals and the beauty of science

A common creationist debating tactic is to sneer at science, saying something like “It changes all of the time! Scientists can never make up their mind, and often times they’re wrong! Why would you want to trust something that admits it could be wrong?”
And my response is usually to laugh, because that’s precisely what makes science so wonderful. We don’t stick with some dogmatic book even when faced with mountains of contrary evidence. We’re constantly trying to figure out where we’re wrong, so we inch closer and closer to an understanding of reality that’s based on…well, reality. Finding out we were wrong and correcting that mistake is the beauty of science.

I bring this up because a recent news story illustrates this perfectly to me. You may have seen the story circulating that non-African humans are part Neanderthal. Yes, some Neanderthals and Homo sapiens interbred back in the day!

At first I was a little confused, because I thought we had established this in May of 2010 when the Neanderthal genome paper by Svante Paabo’s group came out. But this new paper serves as a confirmation of that work, since it avoids one of the main criticisms of the study – that the human and Neanderthal DNA were cross contaminating each other. This new research only looked at human DNA, and compared it to the Neanderthal sequence. What they found was that about 9% of the X chromosome has a Neanderthal origin in non-African humans.

But if I go back to just April of 2010, everything was different. I was taking my 500 level Evolution class at Purdue, about to graduate. Our final project included downloading mitochondrial DNA sequences of humans, Neanderthals, and other apes to determine if humans and Neanderthals had interbred. From that data alone, the conclusion was an obvious “no.” And that’s what all prior knowledge had said up until that point.

I remember one of the last questions on the project being to explain how new information could potentially change this viewpoint. We needed the whole genome before we could definitively say Neanderthals and humans didn’t interbreed! Mitochondrial DNA is only a tiny part of the whole genome. We need more information because we’re so closely related. And what if only Neanderthal males were the ones mating with humans? Then no Neanderthal mitochondrial DNA would be passed on at all!

One year later, and my professor has to totally redo his lesson plans.

And that’s what makes science awesome.

This is post 3 of 49 of Blogathon. Pledge a donation to the Secular Student Alliance here.

Attention Seattle science fans!

The Department of Genome Sciences at UW (aka, mine!) is starting its summer public lecture series, Wednesdays at the Genome. Tonight is the first talk on “Recent adventures in human evolution” by Dr. Josh Akey. I did one of my lab rotations with Josh, and I can assure you it’ll be an interesting, fun presentation. Here’s some more info:

The UW Department of Genome Sciences played an important role in determining the sequence of the 3 billion letters of DNA specifying all of our hereditary information and is now one of the leading centers where the human genome is being interpreted and where new technologies for this analysis are being developed.

To share these advances with the public the Department of Genome Sciences hosts a ‘Wednesday Evenings at the Genome’ public lecture series each summer. These exciting discussions assume no background knowledge in genetics or other biological subjects and provide opportunities to chat with our presenters.

Presentations begin at 7:00 pm in the W.H. Foege Building Auditorium (S060) and will be followed by refreshments at 8:00 pm just outside the auditorium.

ADMISSION is free and the public is especially encouraged to attend!


• July 13 – Mike Bamshad – Confessions of the genome: solving rare disease mysteries

• July 20 – Elhanan Borenstein – Meet your tenants: A genomic tour of your inner microbial zoo

• July 27 – Harmit Malik – Paleovirology: ghosts and gifts from ancient infections

Hope you enjoy it!

I just bought my genome!

Well, kinda sorta. 23andMe, one of the more popular personalized genomics companies, is having a DNA Day sale today. Usually the price to get your genome analyzed (more on this in a bit) is $199 for the kit and $9 a month for a year for their update service – where they’ll rerun your data when new research comes out. But today you can get the kit for free!

I’ve been wanting to do this for a long time but was prohibited by the price, but this is a deal I can’t pass up – so my kit is ordered. I’m prepared to muster up a lot of saliva and then still have some left over to drool over the data. Yep, not only do they give you general interpretations, but you can access the raw data – something a geneticist like me can actually have a lot of fun fiddling around with.

But before everyone runs off and buys their own kit…a warning. I honestly don’t think I’d recommend 23andMe (or any other type of personalized genomics) to a non-geneticist. Not yet, at least. There are a couple reasons:

1. The technology in this area is greatly improving. They just upgraded from a 550,000 single nucleotide polymorphism (SNP) chip to the 1 million SNP chip. That means they’re looking at a million sites in your genome that are known to be variable across humans. While that may seem like a lot, it’s really just the tip of the iceberg. Pretty soon you’ll be able to have your whole genome sequenced. You may want to wait to get the biggest bang for your buck.

2. Genome Wide Association Studies (GWAS) sort of suck, and that’s what a huge chunk of their data, especially the medical stuff, is based off of. GWAS look for SNPs that are associated with a trait, usually disease. The thing is, usually an association can explain a tiny percent of cases of that disease – something like 1%, or even less. And often times that SNP doesn’t always produce a certain trait – for example, having the infamous BRCA gene doesn’t mean you’ll get breast cancer for sure. And almost all studies are done with people of European ancestry, so if you’re not, your results will likely be very inaccurate. So tl;dr, it’s really wishy washy.

3. Because of that, you need to take your results with a grain of salt – which is hard for people without genetic or statistical backgrounds. And that can result in a lot of self-diagnosing that really can’t replace just going and talking to your doctor and giving them a medical history.

People ask if I’m afraid I’ll find out something I don’t want to know – but I’m the type of person who rather know. I’m going to be honest – If I’m predisposed to some horrible disease that will kill me in my 30s, I would not be sitting in a laboratory pipetting or programming. I very much have the view that I want to live life to it’s fullest, and I want to know if I have significantly less years to do so. That and I think learning more about my biology and my ancestry is worth the risk. I’m a scientist and a skeptic – what’s more interesting than the truth?

I obviously won’t share all of my data since much will be very personal, but definitely expect more blog posts about the subject in the future.

You know what else has unique human DNA like a fertilized egg?


Just sayin’.

Science aside of the day:
Well, and T lymphocytes. “T cells” are a type of white blood cell involved in the immune response. They’re special because they undergo something called somatic recombination.

Try to remember back to high school biology. During meiosis (the formation of gametes) there’s a step where Chromosome 1 from Mom and Chromosome 1 from dad can swap chunks of DNA – that’s recombination. It’s the reason why we have so much diversity – because you’re not just getting Grandma or Grandpa’s chromosome, you’re getting a mix of both.

Usually this only happens when making sperm or eggs, but there’s one time it occurs in non-gamete (somatic) cells – in the production of T cells. A protein called a T cell receptor recognizes antigens (foreign particles) from viruses, bacteria, parasites, and even tumor cells. But think of it – if there was just one gene coding for a T cell receptor, we’d only be able to recognize one type of antigen. That’s no good – we need to be able to recognize millions!

Thankfully evolution has the answer. The T cell receptor gene has three main segments: Variable, Diverse, and Joining. There are 65 V, 27 D, and 6 J – but the cell only needs one of each! That’s where somatic recombination comes in – it randomly deletes all but on of each segment, leaving the cell with a unique combination.

“But wait,” cry my more mathematical readers, “that only leaves 10,530 combinations! That’s not very diverse at all!” You’re right! These huge structural differences make up most of the diversity, but these genes are also hypermutable – they gain mutations WAY faster than other genes. So that contributes to the diversity even more!

So, are we ready to start calling every T cell a person because it has a unique human genome? I’m not sure if my psyche can stand all the funeral’s I’d be having every time I get sick.

Genetics will not be used to abort straights OR gays

Not because of ethics, necessarily, but because of science.

Genetics is complicated. This is a concept that all non-scientists, regardless of political leaning, seem to have a hard time grasping. I’ve heard liberals who are worried that advances in genomics will result in a simple prenatal test, which bigots would gobble up to make sure they’re not growing the next Ricky Martin or Ellen DeGeneres. This always seemed like a silly fear, since people from the religious right also tend to be not so fond of abortion.

But it’s not just the liberals. Now World Net Daily is worried gays are going to abort straight babies:

If two homosexual men want to use in vitro fertilization to conceive a baby and then use genetics technology to ensure the baby is also “gay,” while disposing of any “straight” embryos, would the law have any ethical problems with that? John A. Robertson of the University of Texas Law School is the chair of the Ethics Committee of the American Society for Reproductive Medicine and an advocate of what his book “Children of Choice” calls “procreative liberty.” In a paper for the Washington, D.C., think tank Brookings Institution, Robertson presents a futuristic scenario where advancing science and society’s evolving morality could create a once only dreamed-of ethical dilemma:

“Larry, a pediatrician, and David, a wills lawyer, meet in their late 20s, fall in love, and marry on June 15, 2025, in Indianapolis,” Robertson writes. “By 2030, they are well-enough established in their careers to think about having their own child. Larry’s 24-year-old sister Marge has agreed to donate her eggs, and David will provide the sperm, so that each partner will have a genetic connection with the child. … In the process, Larry and David come to realize that they would prefer to have a male child that shares their sexual orientation.” He continues, “The clinic doctors are experts in embryo screening and alteration, but cannot guarantee that the resulting embryos will in fact turn out to be homosexual. To increase the certainty, they will insert additional ‘gay gene’ sequences in the embryos.”

Of course gays, what with their agenda and all, are going to engineer some gaybies! So much more reasonable. Heterosexuals are doomed.

I don’t think you should chose an embryo based on sexual orientation, but let’s put ethics aside for a moment and talk about the science. The ethics debate is irrelevant because the “science” they discuss is ludicrous. As someone who’s studying the “mushrooming” field of genomics, let me try to explain.

Homosexuality almost certainly has a genetic component (1) and has potentially been associated with certain areas of the human genome (2). However, “genetic component” does not equal “gene.” Genetics is way more complicated than what you learned back in middle school – it’s not just single genes with dominant and recessive alleles. You can have multiple genes affecting the same trait, numerous alleles per gene, and interactions between certain combinations of certain alleles between different genes.

If you do find a single mutation that’s associated with homosexuality, it’s likely to be very very rare in humans. If it was more common, we would have identified it a long time ago using traditional genetic tools. Such a mutation would be able to explain just a small percentage of homosexuality. I’m sure by now you’ve heard of studies in the news that have claimed to find a genetic component to heart disease, or schizophrenia, or something. If you read the fine print, it usually only explains something like 3% of the disease. Not really predictive enough to start aborting the breeders.

If this sounds complicated already, it’s just the tip of the iceberg. You can also have mutations in regulatory regions of genes. These aren’t DNA sequences that code for the actual protein, but rather regulate things like how often or in what tissue that protein is made. You can also have copy number variants (CNVs), where some people have extra (or less) copies of a certain gene.

Thanks to massive advances in technology, we can study stuff like mutations, regulatory regions, and CNVs pretty well now… but they’re still not the full story. Often times a single “hit” – one mutation in a gene, or one big deletion in a chromosome – isn’t enough to actually cause a trait. This is especially true when dealing with neurological traits like autism or learning disabilities, and may be implicated in a behavioral trait like homosexuality. Often times you need multiple mutations or deletions – or a combination of both – until you actually show the trait in question.

But it’s still even more complicated than that. It’s not as easy as saying Mutation A + Deletion 2 = FABULOUS! Both of these events are extremely rare, and there are likely thousands and thousands of different combinations of “lesions” (messed up DNA) that could cause a trait. So even if you sequenced a baby’s full genome, you’d have no idea what all the de novo (new) mutations and deletions would do, because they’ve likely never been seen in that combination before.

And all of this isn’t even taking into account epigenetics (which can further regulate DNA, and can even differ between twins), and environmental factors (which can range from hormones you’re exposed to in the womb, to listening to too many show tunes as a small child).

So the odds of Teh Gay being boiled down to a simple test, or a simple gene you can use to infect the population? Basically zero.

Genetics is complicated, and I don’t expect everyone to be able to understand it in depth. Even as a first year PhD student, I tried my best to write the above paragraphs jargon free and without unnecessary detail. But at the very least, admit that it’s complicated and you have no real idea how it works instead of concocting conspiracy theories.

Though I have to admit, it’s amusing that these are the same type of people who claim that simply knowing gays exist, or worse, allowing them to be parents is enough to turn someone gay. Which is it, nature or nurture? Oh right, whatever currently fuels your paranoid hate speech the most.

But you know, maybe the totally wacked out religious types would be content aborting fetuses if they even had a 5% higher chance of being gay. In which case, I’d like to point them to a study that showed each older brother a man has increases his probability of being homosexual by 28% to 48% (3). If they really want to avoid bringing gay men into the world, stop giving birth to sons. And if you’re not willing to rely on abortion, only have one child.

A win-win situation, if I do say so myself.

1. Bailey JM and Pillard RC (1991). A genetic study of male sexual orientation. Archives of General Psychiatry, 48:1089-1096.
2. Mustanski BS, et al. (2005) A genomewide scan of male sexual orientation. Human Genetics, 116(4):272-8.
3. Blanchard R (1997) Birth order and sibling sex ratio in homosexual versus heterosexual males and females. Annual Review of Sexual Research, 8:27-67.

More quotes from the lab

There’s another first year graduate student rotating in the same lab that I’m rotating in, though he’s working on a different project from me. How do our projects differ, you ask?

1st Year: *talking to another labmate about something completely off topic*
Post doc: Hey, that’s five minutes you just wasted that could have gone toward curing autism!
Me: That’s why I’m not studying autism.
Post doc: *laughs* So you can waste as much time as you like?
Me: Yep. Evolution’s not going anywhere!

Joking aside, I actually have been getting a lot of work done. For the fellow biologists: I run my first microarray on Tuesday! For the non-biologists: I get to do cool nerdy stuff I haven’t done before!

This is why I don’t consider myself a science blogger. Too lazy.

Grad school is hard

Obvious statement of the day, I know.

But grad school is also pretty cool. The new quarter has started, and here are my classes:

Advanced Genetic Analysis (first half) – basically how to set up experiments using a bazillion different genetic tricks in order to investigate, well, anything. You know how cool it was solving Punnett square problems? Yeah, it’s like that on steroids. …What do you mean Punnett squares aren’t cool?

Molecular Population Genetics and Evolution (second half)- I can’t wait for this class. Should rename it “Jen has a giant nerdgasm every Tuesday and Thursday.”

Introduction to Statistical and Computational Genomics – I know the title sounds scary, but this will likely be my easiest class. Half of the time is learning how to program in Python, which I pretty much already know. Probably won’t learn anything new until the last couple weeks, where we talk about classes. But the other half of the class is a lecture on bioinformatics, which I basically know nothing about, so that’ll be useful.

My lab rotation still is about human population genetics and evolution, but this time instead of single nucleotide polymorphisms (SNPs) I’m looking at copy number variants (CNVs). …If I was a good science blogger I would take the time to explain what those are, but I have to run to class. Sorry, you’re stuck with Wikipedia for now!