While many of the 23andMe scientists have purified DNA more times than we’d like to remember, there are a fair number of people here (on the science team and on the engineering and business teams) who’ve never spent any time at the lab bench. We love all things DNA around here, so we recently set out to make sure everyone at 23andMe has gotten up close and personal with some Gs, Cs, Ts, and As.

We didn’t use human DNA – instead we isolated the genetic material of strawberries (click here for instructions). It’s pretty simple: smash up some strawberries with some soap and salt, add some rubbing alcohol, and voila! DNA!

As you can see in the pictures and video below, it’s an activity that is fun for all ages!

1 2 3 4 5 6 7

dp

ShareThis

Neanderthal and Homo sapiens skeletons side by side. The thicker femurs, different eye sockets and barrel-shaped chest of our distant relatives are prominent in this comparison.

Mining the past: The Neanderthal Genome Project
The first invited speaker at the SMBE 2008 conference was Svante Pääbo of the Max Planck Institute for Anthropology in Germany. Pääbo and colleagues continue their incredible project to sequence the Neanderthal genome. Neanderthals are especially interesting in understanding our own history; they were another animal that walked upright, hunted with weapons, used clothes, and had culture, traits we consider very “human.” Pääbo presented some new findings that may change the way we think about our own history and that of our distant cousins, who went extinct around 25,000 years ago.

So far, the project has sequenced more than 3 billion Neanderthal DNA base pairs. The figure sounds impressive, and it is. However, sequencing ancient DNA is subject to contamination and in fact more than 99% of the DNA Paabo’s group extracts from Neanderthal bones is from bacteria, fungi or other organisms – including modern humans.

Scientists have debated for decades whether Neanderthals and humans interbred. So far, the Neanderthal genome does not show any evidence of having human ancestry. But the recent split between humans and Neanderthals has resulted in some sharing of genetic material between the species. That is, some people may share versions of SNPs with Neanderthals, but this sharing traces to a common ancestor who lived before the two species split about 800,000 years ago.

One especially interesting finding by Paabo’s group was in the so-called “language gene,” FOXP2. Humans have a very different version of FOXP2 than most other mammals, birds, and reptiles. Rare deletions in the gene cause people to have trouble with speaking and comprehension, providing support that the gene is important for language. Interestingly, other verbal mammals also have changes in FOXP2.
Scientists had thought the “human” version of FOXP2 arose within the last 200,000 years, since the origin of Homo sapiens and long after the human lineage split from Neanderthals. However, it turns out Neanderthals share the human version of FOXP2. These results indicate that something else happened in human history to make FOXP2 appear younger than it really is; and that this may not be related to the unique version of the gene shared by humans and Neanderthals.
So, is FOXP2 the gene that makes us unique from other animals? No. But could it still have played an important part in our own history? Probably. Just one of the many mysteries that evolutionary geneticists hope to answer.

ShareThis

You may have already read about 23andWe and the ”Power of We” in earlier blog posts. As the research arm of 23andMe, we’re hoping 23andWe can produce valuable discoveries about the genetic roots of diseases, conditions and traits that are little-studied due to funding limitations, logistical obstacles or simple lack of interest among scientists. 23andMe’s payoff could prove enormous by offering insights that may eventually lead to risk predictions, diagnostics, treatments or even cures for diseases.

At 23andMe, we are all about grand ideas with big potential, and we are committed to realizing our vision, but how do you even start such a big project? You begin at the beginning, of course.

When we first started thinking about the traits we wanted to study in 23andWe, we faced a big problem. Out of all the possible traits out there, which ones should we study first? It’s both an overwhelming and wonderful problem to have, and one, I think, that is pretty unique in the scientific community.

Most academic human genetics labs don’t have the luxury of splitting their research resources amongst many different problems. Subject recruitment, genotyping, and analysis are so difficult and expensive for even one trait that it’s usually feasible for only the largest and most well-funded centers to do this type of work. Limited funding usually also means that only the traits and diseases that are considered really serious get studied.

In 23andWe, we saw a unique opportunity to tackle interesting and important questions about biology and disease that have so far been largely left unanswered.

Focusing on what has traditionally been understudied helped narrow down our field of research questions a little bit, but not that much. Frankly, not that many traits and diseases have been studied very well. So we had to apply a few more filters.

We wanted to pick topics that
• Are relatively simple, because we have to walk before we can run.
• Are easy to capture accurately over the Internet (which for now will involve online surveys).
• Will apply to a wide swath of the population.
• Already have some evidence in the literature for having a genetic basis

I can’t tell you how many PubMed (a biomedical literature search engine sponsored by the National Institutes of Health) searches we’ve done, nor how many 40-, 60-, and 80-year old papers we had to dig up to find some of the information we needed.

It’s been a fascinating project to work on. Did you know that people have been studying the inheritance of tongue curling for close to 70 years ? Turns out it’s not the simple Mendelian trait your high school biology teacher told you it was. Instead, tongue curling ability is probably determined by a mix of genetic and environmental factors.

It’s not all about tongue curling though. We’ve also been delving into the literature on many more serious health issues that affect large segments of the population but have not yet received as much attention as they deserve. These topics range from dyslexia, to endometriosis, to migraine. We plan to collaborate with experts in the scientific and medical communities to push forward the limits of our knowledge on these important topics, with the long-term goal of helping to improve the quality of people’s lives.

In the coming months you’ll start seeing the fruits of our labor as we roll out our first 23andWe projects. Of course, this is just the beginning. We’re going to spend countless more hours scanning our PubMed search results and fighting with the scientific journals’ user-unfriendly websites. And our thought processes, methods, and goals will continue to evolve as we gain experience and accumulate data. But it’s all worth it, because the rewards are potentially great.

Stay posted…

ShareThis

Art Nouveau architecture at the Codorníu winery outside of Barcelona.

So much new research was discussed in Barcelona it’s hard to decide which were the most notable presentations. But here are a few of the ones I found most interesting:
Were humans shaped more by history or local environment?
A major debate in the human evolutionary genetics talks and posters considered the origin of the genetic differences seen in humanity today: Were they shaped more by populations splitting apart and coming together, or evolutionary adaptation to local environments? Interestingly, people from the lab of our SAB member Jonathan Pritchard presented arguments on both sides. Both talks presented strong evidence using similar data sets. Perhaps one phenomenon has more impact locally and the other more regionally. Certainly the debate continues.

James Cai and coauthors from Stanford (including our very own R&D scientist Mike Macpherson) and The Hebrew University of Jerusalem showed that the history of the human genome cannot be explained simply by neutral variants – variants that do not cause a functional change. All across the genome there is evidence of “selective sweeps” where an advantageous version of a gene quickly increased in frequency in a population or species. For example, the gene FOXP2 has undergone a selective sweep in all humans within the past several hundred thousand years and may have contributed to our ability to use advanced language. More recent selective sweeps in the Duffy and Lactase genes (both have variants that 23andMe customers or demo account holders can read more about in My Gene Journal) happened after human populations diverged and thus didn’t sweep across the entire globe but are confined to specific regions: primarily western Africa for the Duffy-0 variant and Europe, the Near East, eastern Africa, and southern Asia for Lactose Tolerance.

Selective sweeps tend leave evidence in the form of nearby DNA that gets dragged along with the variant as it sweeps across a population. Similarly, new variants that are disadvantageous (or become disadvantageous when, say, moving into a new environment) can leave these similar signals as they are dragged out of the population. However, it is often difficult to separate out effects of population history from these selective forces. By using a novel statistic that controls for population history, Cai and colleagues show that many locations on the human genome have been affected by these selective sweeps. While previous scans for positive selection required these selective sweeps to be incomplete (see here and here, for example), the authors use a metric which can go back even further to look at the timing and strength of selective sweeps which have affected the entire human population, even going back as far as one million years. This work is an extension of previous research on Drosophila.

Interestingly, one of the data sets used for this work was the complete genome of Jim Watson, who co-discovered the structure of DNA.

Population Structure, History, and Migrations
Sarah Tishkoff of U. Penn gave a talk on her incredible data set of sub-Saharan African populations. So much of the world’s genetic diversity is located in this region, yet its inhabitants have been relatively under-sampled so far. Tishkoff’s data, in the context of global variation, makes it apparent just how important it is to understand the history of sub-Saharan populations in order to understand the history of our species. In one example, Tishkoff used a technique known as Principal Components Analysis (PCA) to collapse all their genetic data into three dimensions. Individuals near each other in PCA are more similar. In her plot, a hunter-gatherer population from Tanzania known as the Hadza can be found in their own dimension on the plot, which suggests that the Hadza, while having a small population size, have been isolated for a long, long time and are quite divergent from other populations, even including the 52 in the CEPH-HGDP data.
Tishkoff also showed how difficult it is to extrapolate from one African population to the next, even if they neighbor each other. One example of this is in parts of western Africa where the Fulani have increased malaria resistance compared to other groups such as the Mossi and Rimaibe – even within the same town.
Several talks and posters looked at the new lactase persistence variants discovered last year in sub-Saharan Africa and the Near East. These variants are functionally the same as their much more common counterparts, which allows Europeans and South Asians to drink milk into adulthood without experiencing lactose intolerance (23andMe customers can look up their genotype for this variant in My Gene Journal). But because they differ genetically, these newly discovered variants illustrate the importance of milk digestion for populations that relied on herding in their past. Multiple research groups showed that the eastern African persistence variants made their way down to the San Bushmen and neighboring populations of southern Africa.
When normal inheritance breaks down

Genomic imprinting in action. Here, the color of the offspring comes from the father, regardless of which genotype he has.

Andrew Clark of Cornell has been looking at versions of genes in mice that change the traits of offspring depending on whether they are inherited from the mother or father. This phenomenon, called Genomic Imprinting, has been detected in many mammals before, including humans, although interestingly it isn’t found in marsupials or the egg-laying monotremes like the Platypus. However, the traits affected by genomic imprinting have not been surveyed using a genome-wide approach.
Clark and colleagues used the Solexa sequencing platform to look for differences in the mouse brain between mice crossed from two different strains. By switching the strains of the mother and father researchers can detect traits that derive exclusively, or “imprint on”, one parent.
It turns out a good number of genes exhibit genomic imprinting Genes imprinted on the father tend to show only the trait of the father. Genes imprinted on the mother tend to let some of the father’s trait come through, albeit at much lower numbers. In addition, the researchers found differences in the organs affected by imprinting: genes imprinted on the mother were more likely to be expressed in the reproductive organs and those imprinted on the father were found more in the brain.
It appears that imprinting has no immediate benefit for offspring and may have originated in mammals completely by accident, a quirk of our histories. But learning about how imprinting evolved will help us understand how they came to be.

ShareThis

A large international collaboration of researchers has tripled the number of SNPs associated with Crohn’s disease, an inflammatory bowel disease that affects about 500,000 Americans. Their results were published online Sunday in Nature Genetics.

Data from more than 16,000 people with European ancestry – 8,059 from a combination of three previously published reports, a new set of 2,325 patients and 1,809 control subjects, and a group of 1,339 families – identified 21 new SNPs associated with Crohn’s disease and confirmed 11 previously implicated SNPs.

(See the table at the end of this post for SNP details).

Although many of the SNPs described in the report only slightly impact the risk of developing Crohn’s, the findings could be important in helping researchers understand the basic biology of the disease.

“Studies such as this are not about developing diagnostic tests, but about identifying targets for new drug therapies. Crohn’s disease can be a very serious condition, often requiring surgery, and the sooner we can understand the underlying causes, the sooner we will be able to devise new treatments to help our patients,” said study author Dr. Miles Parkes.

Some of the newly identified SNPs are in (or near) genes linked to diseases other than Crohn’s, including type 1 diabetes, arthritis, psoriasis, and asthma. This highlights “the important relationships between different diseases and, as such, may offer valuable insights into the pathways that lead to common symptoms such as inflammation,” said Dr. Mark Walport, director of the Wellcome Trust, one of the funding agencies that supported the study.

The researchers suspect there are many more genetic associations left to find for Crohn’s disease.

“These [SNPs] explain only about a fifth of the genetic risk, which implies that there may be hundreds of genes implicated in the disease, each increasing susceptibility by a small amount,” said the study’s lead author, Jeffrey Barrett.

23andMe customers can check their data for many of the SNPs found by Barrett et al. using Browse Raw Data. Here we list 23 out of the 32 SNPs detailed in the report –six SNPs that are already used in the My Gene Journal Crohn’s disease article have been omitted, as have three SNPS that are not currently available from 23andMe’s service.

Out of a possible total of 46, 80% of people will have between 17 and 25 copies of the riskier versions of these SNPs. Because there are so many of these SNPs – and because the riskier versions are often more common than their less risky counterparts – even people at the high end of the distribution are not much more likely than average to develop Crohn’s disease.

“Effect” is the increase in odds of developing Crohn’s disease conferred by each copy of the riskier version of a SNP. SNPs where we are providing a proxy for the SNP originally reported in the paper are marked with an asterisk.

ShareThis

The Stormtroopers in Star Wars were modeled after these air vents at La Pedrera in Barcelona.

Part One: Benvinguts a Barcelona!

The pace of genetics research has increased dramatically over the past few years. What was possible only in a large, well-funded lab a few years ago can now be done by a solitary grad student on a laptop.
Many people at the conference were studying large publicly available sets of genetic data, such as the 23andMe-sponsored data set of 650,000 SNPs from 1,000 individuals in 52 populations (data available here, for those interested). Others were taking advantage of next-generation sequencing platforms (such as 454 and Solexa) to investigate everything from differences in protein expression in different strains of mice to the genetic makeup of extinct organisms – even Neanderthals.
Over the next few posts in this series I’ll discuss some of the most interesting talks and topics; but I’ll start with why we went to Barcelona.

Maternal history of populations

We were happy to present research we’ve done that puts a date on every major branch point in the human mitochondrial DNA tree. Because mitochondrial DNA is passed directly from mothers to their offspring, it can be used to trace the maternal ancestry of every person on the planet (for customers and demo account holders, that’s the maternal line feature). We used more than 4,000 publicly available complete mitochondrial genomes from Genbank, assigned to maternal haplogroups using their full sequence, to accomplish this goal.
Why would we want to do this? Well, calculating the dates of common ancestors allows us to tell someone how long ago he or she shared a maternal relative with a friend, coworker, or even a celebrity. Our study is the first time all the haplogroups on the tree (or at least over 550 of them) have been dated all at once. For example, thanks to our research 23andMe can state confidently that Jesse James and Jimmy Buffett can both trace their female lineages back to a single woman who lived 60,000 years ago, probably somewhere in the Near East.

In addition, by looking at dates of common ancestors across the entire maternal tree we can fill in some details relating to the history of our species. We know that everyone living on Earth today descends from a woman who lived in Africa around 175,000 years ago. But each lineage that connects back to that woman has had a different history of mutations. Looking at this mutation history can tell us about the different groups of humans that expanded from eastern Africa to settle across the planet. For instance, we might be able to tell if certain groups were large or small when they split off from the tree, or whether they faced different environmental challenges that led to their rise or fall.
Interestingly, we see an increase in mutation rates along certain mitochondrial DNA lineages that arose after the out-of-Africa expansion around 50-60,000 years ago, but before the glaciers retreated in the last Ice Age (around 18,000 years ago). This contradicts previous work – maybe because we used much more data than previous studies and it was gathered from a much more diverse set of people. Whether the increased mutation rates were due to natural selection as humans moved into different environments or other events in the history of our ancestors remains to be seen – we’re working towards resolving the importance of various evolutionary processes in the history of the human maternal lineage.
In the next installment, I’ll blog some more about other interesting topics and presentations at the SMBE conference.

Adéu for now!

Click here for a PDF of our poster entitled: “How do you date all humans at once? The use of complete genomes to date nodes on the human mitochondrial tree.”

ShareThis

Since right-handers outnumber southpaws by approximately 9 to 1, it’s not hard to imagine why there’s a bias against lefties. Yet there are plenty of left-handed role models – two of the last three presidents (and both of this year’s contenders) favor the left manually, if not necessarily politically. Celluloid hero Luke Skywalker (Mark Hamill) dispatched the Death Star with his left, while Neo (Keanu Reeves) favored his sinister side in harnessing the power of the Matrix. And after granting a mortal man divine powers, guess which hand God (Morgan Freeman) used to snatch them back?

Which hand do you use to hold your coffee cup? Is it the same hand you write with?.

Though most people define left- or right-handedness by the hand that grips the pen, it turns out there are varying degrees of handedness. Want to find out if you should actually consider yourself either purely left- or right-handed, or somewhere in between? Take the new 23andWe survey on handedness and find out.

The non-right-handed 10 percent of the population has proved useful to scientists who want to understand how the brain works. In her “right-shift gene” theory, for example, British psychologist Marian Annett argues that handedness is merely a side effect of having a single (unidentified) gene evolve to assign speech control to the brain’s left side (which governs movement in the body’s right side). Her proposal is based on a study of hand preference when performing various tasks, and then resulting subgroup classifications. Analyzing the distribution, she noticed that there was a tendency toward right-handedness even among some lefties, which led to the name of her theory. If the gene is damaged early in development, she says, it would affect the left hemisphere of the brain and could be responsible for conditions such as autism and schizophrenia.

For Australian psychologist Michael Corballis however, handedness might have been part of what helped separate man from monkeys. He thinks language and right-handedness evolved together, and ties this relationship in to the asymmetry of the brain. Most people, Corballis notes, are right-handed and have their language center in the left hemisphere of the brain. He posits that they have two copies of a “right-handed” gene. Corballis says that lefties, however, have one copy of the right-handed gene and what he calls a “chance” gene, so their language center may be in the left hemisphere, the right hemisphere, or even in both. The disruption of brain asymmetry, he says, explains the link between left-handers and language problems such as dyslexia and stuttering.

Both psychologists’ ideas stress the importance of understanding how the brain is organized, and looking at brain development on a genetic level may give researchers an even better understanding of the subject. Though genetic evidence to support either Annett’s or Corballis’ theories has yet to be found, last year researchers at Oxford’s Wellcome Trust Center for Human Genetics announced that they’d found a gene that is associated with left-handedness if children inherited their father’s copy. Known as LRRTM1, the gene has also been associated with dyslexia and schizophrenia.

Though LRRTM1 is the only gene that has been associated with left-handedness to date, plenty of other interesting traits have been linked to it that also offer hints to its biology. Consider these other findings that, as yet, haven’t been linked to a gene or genes:

• there are more left-handers in England than there are in Wales and Scotland combined;
• there’s a link between studying architecture and left-handedness …
• … and between creativity in painting and music and left-handedness;
• the percentage of left-handers among homosexuals is higher than in the general population;
• an increased incidence of breast cancer among post-menopausal left-handed women in Australia and America; and,
• slightly more left-handed major league baseball players were born in June than in any other month.

Some of the findings above may seem trivial, but some are also controversial. One study that’s been hotly contested for the last 20 years, for example, is whether or not right-handers live longer than left-handers. Several researchers have since published reports that either support or contradict the data.

Images from: bigfoto.com and PLoS ONE: Kalisch et al, 2006.

ShareThis

Bertalan Mesko of scienceroll.com kindly arranged for us (ErinC and joyce) to give a presentation about our company on Second Nature, an island operated in Second Life by the Nature Publishing Group. We talked about the basics of our service, and our newest addition – 23andWe.

It was a great – if somewhat weird — experience. Neither of us has ever given a talk where a horned blue monster and a robot made out of boxes were in attendance!

Thanks to everyone who came to watch!

If you missed it, Bertalan live-blogged our presentation. Our poster will also remain on Second Nature if you want to check it out.

ShareThis

Aspergillus fumigatus

Many cancer patients who receive high doses of chemotherapy or radiation undergo a bone marrow transplant to replace blood cell-generating tissues that are killed as part of their treatment. But after a transplant these patients face a new challenge – opportunistic infections that take advantage of their weakened immune systems.

One common and life-threatening infection among bone marrow transplant patients is invasive aspergillosis (IA), which is caused by the fungus Aspergillus fumigatus. About 10% of bone marrow transplant recipients are affected by invasive aspergillosis (IA). An estimated 30% of those affected will die within three months.

A study published in the June issue of PLoS Genetics shows that a SNP in the gene encoding a protein called plasminogen significantly affects the risk of developing IA after a bone marrow transplant. Compared to transplant patients with the GG genotype at rs4252125, the risk of developing IA was increased 3.0 fold in people with the AG genotype, and 5.6 fold for people with the AA genotype.

Zaas et al began their study of IA in bone marrow transplant patients by looking for clues in several genetically distinct strains of mice. The researchers mimicked a bone marrow transplant in the mice by giving them drugs that suppressed their immune systems. They then exposed the animals to the Aspergillus fungus.

The researchers found that some mouse strains survived Aspergillus infection, while others did not. A SNP in the plasminogen gene correlated with Aspergillus susceptibility, and the researchers considered it a likely candidate because other scientists had recently indicated that proteins in the same biological pathway as plasminogen are important for fighting off infectious pathogens.

Armed with this clue from mice, Zaas et al turned their attention to humans. They followed a group of 230 immunosuppressed bone marrow transplant recipients for one year. Just as they had seen in the mice, the researchers found that IA was associated with a SNP (rs4252125) in the plasminogen gene.

Oddly, the association between rs4252125 and IA applies only 40 days or more post-transplant. The authors aren’t exactly sure why this is the case. It could be due to the low proportion of patients who are infected with Aspergillus in the first 40 days after their transplants. It could also be that until 40 days, when the transplanted bone marrow really starts to take hold, other risk factors swamp out the effect of rs4252125.

Laboratory tests showed that the plasminogen protein encoded by rs4252125, which is made in the liver, binds to Aspergillus fungus. And computer models indicate that the SNPs found in both mice and humans change the protein in a way that alters binding to the fungus, giving a possible explanation for why the SNP affects susceptibility to IA.

Though more research and evaluation is required before clinical applications can be developed based on this research, the authors write in the conclusion of their report that their results have “important implications for pre and post transplant care, and may also have implications for the management of other immune-compromised patients. For example, genetic testing could identify high risk individuals who may benefit from use of broad-spectrum antifungal agents or enhanced monitoring for infection.”

ShareThis

ShareThis