Friday, March 27, 2009

Sharing

One of the most thrilling things you can do with your genetic test results from 23andme is sharing the information with others, in search of unknown relatives. This is how it’s done: when we find someone, for example through a website forum, that we are curious about comparing our genes with, we simply send that person an invitation and wait for an answer.

Almost everybody is willing to share at the “basic” level, that is, as long as no specific DNA details are revealed, namely in terms of disease risks. Even so, it took some effort on my part to start sharing with people I don’t know, but once I understood that my privacy was not at stake, I quickly became addicted. Currently, I already have 17 “friends” in this genetic social network, each and every one of them an absolute stranger, that I invited or who invited me to share.

Once the sharing principle is accepted by both parts, it becomes possible to compare genomes through a functionality called “family inheritance”, that allows both persons to see if there are coinciding pieces of DNA in their respective chromosomes.

In fact, this option is supposed to be used to compare genes among family members, who know from the start that they share big chunks of DNA. And, from what I’ve read, it’s pretty unlikely to find this type of genetic coincidence, by chance, in people we’ve never heard of before.

So imagine my amazement when I found out that 4 out of my 17 “friends” are probably distant relatives, distant “cousins” of mine! It’s hard to believe, but that’s how it is: when I compared my genome-wide genetic data with the data from those 4 people, I discovered that whole segments of certain chromosomes suddenly turned blue on the diagram. In one case, the coincidence spans a sequence of 20 million DNA “letters”, or bases, on chromosome 5. In another, as the image shows, the coincidence involves a 10 million-base sequence on chromosome 4 (I’ve hidden the person’s name for obvious privacy reasons).


What does this mean? It means that, according to the partial genetic results we both received, we both inherited, from one of our parents (either our mother or our father), an identical chunk of some chromosome. In other words, we our DNA present a so-called “half-identical” match (i.e., contributed by only one of our parents). Which, in turn, means that our mother and the other person’s father (or whichever combination is responsible for the coincidence) shared (still share) half-identical chunks of their own DNA too. That is, they also were (are) distant (though slightly less distant) “cousins”. And so on, right back to a common ancestor who probably lived a few centuries ago.

Here is the answer I got from another one of my “cousins”, with whom I share a half-identical DNA sequence 10 million-bases-long on chromosome 9, when I voiced my doubts about such a conclusion: “Yes, it means that we definitely share a common ancestor. The precise point in time may be 300-400 years ago, depending on the rate of recombination in that region. It is unlikely that the common ancestor would be further back than 300-400 years, because recombination would have eliminated it from our common ancestry.” Seems reasonable enough to me.

I also asked a scientist I know who specializes in genetic ancestry and she confirmed.

Apparently, such a state of affairs is not that uncommon in those human communities, namely religious, where people didn’t marry much outside of their group – and which, at some point, became part of a Diaspora, spreading worldwide because of political and religious contingencies that arose in their home-countries.

Monday, March 23, 2009

A4793G

This is the codename of the point-mutation, in my mitochondrial DNA, which signals that I belong to the maternal haplogroup (or matrilineal line) known as H7. It means that, at the 4793rd position in my mitochondrial DNA (hightlighted in purple on the image below), instead of having a letter A (the DNA base, or building block, called adenine), my sequence contains another base, namely a guanine (or G). But that’s not the only ancestry-defining point-mutation which is present in my mitochondrial DNA, for it obviously also contains the series of mutations that accumulated, generation after generation, in my maternal ancestors’ genes since the so-called mitochondrial Eve (the mother of all modern human beings, who, experts agree, lived in Africa around 200,000 years ago) walked on this planet.



At 23andme, they tell me that the sequence of the point-mutations which, starting with A4793G (4793 for short) and travelling back in time, go from haplogroup H7 back to mitochondrial Eve, is as follows: 4793, 2706, 7028, 11719, 12705, 10398, 10873, 15301, 8701, 9540, 1018, 769, 13650, 16278, 3594, 4104, 7256, 7521, 10810, 15301, 16129, 16187, 16189, 825, 8655, 2758, 2885, 7146, 8468, 16230, 11914, 10589, 6185, 4312. (A soupful of numbers indeed!) Mutations at positions 2706 and 7028, for instance, define haplogroup H, of which H7 is a direct sub-haplogroup. Thus, going down the tree from the branches to the trunk, one goes through a series of increasingly ancient haplogroups: HV, R, N, L3, L2, L1, and on back to that single, primordial mother.

Monday, March 16, 2009

I want more!

I would really like to have the whole of my mitochondrial DNA sequenced (this is the part of our genome that tells us about our direct matrilineal descent). The best website for this seems to be Family Tree DNA. But out of the three sequencing options they offer – mtDNA, mtDNAPlus, and mtFullSequence –, the only one I’m interested in, because it would give me more information than what I’ve already got from 23andme, is the third one – and currently, that costs 495 dollars. I’m still pondering.

Both the other options would give me much less information than what I already have, and, accordingly, a lower resolution on my maternal “haplogroup” (genetic family). The first option (129 dollars) only reads 22 SNPs (point mutations) in the so-called Hipervariable Region 1 (HVR1), which is like looking at the mtDNA with a simple magnifying glass; the second one (189 dollars), adds to that a series of SNPs from Hipervariable Region 2 (HVR2), which gives a somewhat higher resolution. But compare this to the data from 23andme, which reads some 3,000 SNPs not only from HVR1 and 2, but also along the rest of the whole mtDNA molecule. I rest my case. The advantage of FTDNA, though, is that it’s been around longer and has many more clients than 23andme, which means many more people with whom to compare our results in search of similarities.

Wednesday, March 11, 2009

Chromosome painting

Here is a view of my chromosomes 1 through 22 (actually, each bar represents the two copies of each of my chromosome, one inherited from my father and the other from my mother). There are two more chromosomes (the 23rd pair), which are not drawn here, and which determine the person’s sex. In my case, since I’m a woman, I got an X chromosome from each of my parents and I am XX.

This functionality of the 23andme website, called “ancestry painting”, allows me to see the origin of my chromosomes in terms of large geographical regions.




In my case, the image shows that 99 percent of my chromosomes 1 through 22 are of European origin (they are painted black). The rest of my genome, less than 1 percent, is painted orange and is of Asiatic origin. This encompasses a small fragment on chromosome 8 (inherited from just one of my parents, as seen from the fact that only the lower half of that fragment is painted orange); and a slightly larger fragment on chromosome 18, that seems to have derived from the same geographic origin in both my parents, as it is completely painted orange. Chromosome painting has another very interesting application, which I will explain on some other occasion.

Friday, March 6, 2009

Origins

For the sake of consistency in calculating the estimates of our risk for this or that disease, one of the things we need to define at the 23andme website is our “ethnicity” – which simply means, actually, that we have to specify the geographical origins of those of our ancestors whom we know of. But when I tried to do this for the first time, I didn’t find what I considered to be my appropriate ethnicity, namely Eastern Europe.

My grand-parents all came from the Ukraine, near Odessa, which is clearly Eastern Europe, but I only had the choice between checking a box for Northern Europe or one for Southern Europe. Where was Eastern Europe? Did I have to choose between those two? The closest choice, I seemed to me, was Southern Europe, but that didn’t seem specific enough – not the least because the stories of the Iberian Jews (Sephardic) and the Jews from Central and Eastern Europe (Ashkenazi) are pretty different. The short reply to yet another email to 23andme was: “You can use Southern Europe as the option”. OK; so that’s what I did.

I must say that I was pleased to learn afterwards that my genes also confirmed my own, unlisted, first choice. Using a functionality called “global similarity – advanced view”, I was able to see that the group I was closest to in terms of genetic similarity were the “Ukrainians” – and that they were placed, adequately, between Southern… and Eastern Europe (I’m the larger green plot on the maps below).





Tuesday, March 3, 2009

Things evolve very quickly!

While I was waiting for my results, I had the opportunity to take a look at the kind of results I would be getting. The 23andme website allows you to create a user account right away and to explore the genes of a fictitious family, the Mendels (the name is obviously a discreet homage to Gregor Mendel, considered to be the father of genetics).

One of several questions I asked myself at the time concerned the fact that, apparently, they weren’t going to test for mutations in the BRCA1 and BRCA2 genes. But it so happens that mutations in those genes mean a high risk for breast cancer. I sent them an email asking about it and this was the answer I got: “In the future we hope to include additional analyses that will identify some of the most common breast cancer-associated BRCA mutations.”

The future was actually right there and then, except I didn’t know it until a few weeks later, when I got my results, which included the analysis of three possible mutations in those (in)famous genes. But I didn’t realize that right away. I opened the webpage which contained my results on those genes and only then did I start to grasp the significance of what I was looking at. It was really scary because I was confident – since the future is supposed to lay in the future, not in the present – that those were not the genes I had asked 23andme about. They were. Happily, I don’t carry any of those three mutations.





But the mere fact of not being a carrier doesn’t mean that I don’t carry any other of the countless mutations that can affect these genes. As they had also explained in that reply email, what I got isn’t a diagnostic test of my real risk for hereditary breast cancer. For that, I would have to get the full sequence of those genes, which means they would have to read all the letters in it, and identify every possible mutation. 23andme doesn’t perform this kind of sequencing.

Sunday, March 1, 2009

Journey to my genes

(click on the title to read the whole text)
View PDF of the print story (in Portuguese)


The ideia of learning about the secrets of my DNA has fascinated me for years. Not so much because I want to know about my predisposition to this or that disease, but for what my genes could tell me about my deep ancestors’ origins. Recently, the price of these genetic tests became affordable enough for anyone to have access to them (for a few hundred euros) and I struck the following bargain with my editors at my newspaper: if they paid for the tests, I would write about what they revealed – the good things, the bad things and the so-so things. Here they are. Welcome to me!




The instructions tell me to spit into the little plastic tube. It’s 10 a.m. on a December morning. In the bathroom, I open the test tube and start spitting into it. I haven’t eaten or drunk anything over the last half-hour, I haven’t brushed my teeth either (more instructions). I can’t get this wrong, it has to work without a glitch.

Contrary to what one might expect, it’s not easy to fill with more than an inch of spit a small tube about a half-inch in diameter, while “trying not to make bubbles”, as the instructions go. To facilitate the process, they tell us to wag our tongue against the inside wall of our cheeks. This ensures that together with our saliva, cells from the lining of our mouth will also fall into the test tube. These are the cells that contain the DNA that is going to be analyzed.

- full text -


I do this over and over again. It seems my mouth is getting drier and drier, but it’s just a feeling: the spit slowly fills the tube until it reaches a mark for the desired amount. I’m not really spitting, it’s more like drooling in a very controlled fashion. I look at my watch: I’ve been at it for almost 15 minutes.

I close the tube with a large cap, a special cap that leaks a buffer solution into the saliva sample when you screw it on, to preserve it. I shake the whole thing vigourously, unscrew the large cap and screw on another one, a smaller, normal, final cap. Done. I slip the test tube into a padded envelope, addressed to somewhere in California, that I will dispatch by special mail later in the day.

I ordered this genetic test kit online, in the beginning of December, at the 23andme.com site. A few days later, I received, by special mail, a glossy little green box with my name and surname clearly written on it in capital letters, along with a code number (claim code) that would give me access to my results when they were ready. It was inside the box that I found the test tube, the two caps and the list of instructions.


Genes with a name and a surname

23and me is one of several businesses that currently offer to analize their individual DNA to the public at large. It’s one of the best-known and also one of the most serious. Equally attractive is the fact that they drastically cut their prices at the end of last year. For just 399 dollars, they provide not only a guided tour of our chromosomes, in search of the hidden secrets of our DNA – mostly disease risks (brrrr!) –, but also a description of a series of physical and psychological attributes and even a glimpse of our ancestors. For the first time, it's become possible to obtain information on our own DNA, a DNA with our name and surname, and not just the DNA of an anonymous representative of the Homo sapiens sapiens species.

It doesn’t mean they read our genome’s full sequence – that is still out of monetary reach of the great majority of us. Instead, they only read a (tiny) part of the six billion letters or bases that make up our DNA (half of them inherited from our mother and the other half from our father, joining in pairs on our chromosomes).

This more modest alternative consists in detecting bits which are only one DNA-letter-long, so-called SNPs (single nucleotide polymorphims, pronounced snips), and which may harbor point mutations. When a snip is located in the middle of a gene, it alters its function; but even when that snip doesn’t affect any gene, the fact that is is mutated can suggest that, close by, there is a significant mutation (significant in terms of our health, for instance) on some yet-to-be identified gene. In this latter case, the snip acts as a marker, as a signpost and can contribute to the identification of important genetic mutations. “The SNP is then useful as a landmark, similar to how people describe locations”, explains 23andme on their site. “You may not know where the local hardware store is, but if you know it’s within a block of the pharmacy you went to the other day, you will be able to find it.” It is estimated that the human genome contains about 10 million SNP, and at 23andme they read 550 thousand of those from our cells, wisely distributed along all the chromosomes.

The main aim, says 23andme, is to inform people about their health risks, because that information – which could lead to life-style changes on our part and even make us more alert with respect to certain disease symptoms – is a new form of empowerment, of control we have over our own life.

My motivation, though, was somewhat different: the idea of knowing my DNA, of looking at my A, T, G; C sequences had fascinated me for years. I imagined it would reveal fascinating things about my ancestors. I was never very interested in my recent genealogy, but genetic genealogy, the possibility to discover my roots at the scale of thousands of years, that seemed to me to be worth the trouble. And if in so doing I also managed to link that genetic information to historical data about my more recente ancestors, so much the better.

So that is why, when the opportunity presented itself to become a guinea-pig in these experiments in “personal genomics” (that’s how they call this new field), I couldn’t resist it. Of course, I pondered the risk of discovering terrible things about myself in the way of inheritable diseases, biological deficiencies or the like, against which I would be powerless. Would it be better to know or not to know my risk for breast cancer? But curiosity had the better of me. And that’s how I ended up spitting into that little tube.


Moment of truth

I received my results almost two months later in the form of an email: “Congratulations! Data for the person listed above are now available on the 23andMe website.” And then, a reminder of my login name and a link, and another link for the Getting Started Guide…

I entered my user and my password, ready to face the results without hesitating. But I must confess that I felt somewhat like I had when, some years earlier, I had gotten the results of an HIV test my bank had required for a loan to buy a house: pretty scared, even though I knew the odds of getting a bad verdict were very slim. And so, just as I had done on that occasion, I also asked my husband to stay at my side and hold my hand…

"Welcome to you" was the first thing I read when I accessed my personal page. Underneath that sentence, several sections: My Health and Traits, a survey section, and another one titled “global similarity”. Where were my results? Where were my genes? My list of diseases? My ancestral origins? At first sight, it was all pretty confusing. I started by looking at my “raw data” – that is, at the sequences of my DNA letters that had been decoded –, and I soon discovered that they made, to say the least, for indigestive reading. There were lists and lists of thousands of numbers, letters, gene names, links to incomprehensible genetic databases, comparisons with “reference sequences” (whatever those were) I had never heard about before.

I went back to the initial page and, determined to deal first with the issue of disease, I clicked on the My Health and Traits link. One of the first things I had to do there was to decide if I wanted to see (opt-in) the results concerning a type of a hereditary breast cancer. I was a little nervous with the whole situation, imagining what terrifying, unavoidable things I was about to find out, so I made a hasty decision when I should have pondered more carefully: I happily clicked on the opt-in button, confirming that yes, I wanted to see the results. I quickly scanned the new page to the bottom – and only then did I realize what it was all about.

It so happens that these results concern mutations that can affect to sadly famous genes, called BCRA1 and BCRA2. These genes, although they are responsible for only 2 percent of breast cancers, dramatically increase, in their carriers, the probability to develop breast (and ovary) cancer – to about 50 to 80 percent. At this level of risk, it is difficult to tell it apart from absolute certainty. Moreover, as I read on that same page, in Ashkenazi Jews (a group to which I belong), those mutations are responsible for 80 to 90 percent of hereditary breast and ovary cancers. Luckily, I don’t carry any of the three mutations that 23andme test for. But after the fact, I got cold sweats: what would I have felt if the verdict had been different? How do you live with a thing like that hanging over your head?


Diseases by the dozen

I also found out that I have increased risks (but not certainties, here things are not so harsh) for two other diseases: three times more (1.4 percent) than average to get Crohn Disease (an autoimmune, chronic inflammation of the bowel) and almost twice the risk (2,3 percent) of getting type 1 diabetes (caused by the destruction of insulin-producing cells in the pancreas). The first disease has a hereditary component of 50 to 60 percent; the second one, of 72 to 88 percent. Curiously, as far as I know, I don’t have a family history for either of them. But from now on, I’ll have to be on the lookout for symptoms – and talk to my doctor to see if there’s anything I can do in terms of prevention.

My report contains 102 items – for diseases, serious mutations, physical traits and others. In some cases, the relevance of the SNPs which were analyzed has been largely confirmed by research, but the impact of the mutations is not so strong. In other cases, there is still no scientific consensus on the relevance of the corresponding SNPs. A lot of the information they give us is based on preliminary results. But even so, the results are dully presented and submitted to our consideration. There are links to the most recent scientific papers on each subject – which are not always enlightening, since their conclusions are frequently in contradiction with each other. But there’s no doubt about the fact that it’s all neatly done.

A quick overview: I have more than average risks for a series of cancers, average risks for others, and less than average risks for others. Who doesn’t? Luckily, I don’t seem to be a candidate for multiple sclerosis, a disease for which the genetic component is rather strong (24 to 86 percent).

My risk for having a heart attack (which may have a hereditary component of up to 52 percent) is average. Needless to say, having an “average risk” can be pretty uncomfortable all the same, if that average risk is high… And, for heart attacks, it’s around 20 percent. Even worse is the fact that for obesity, where my risk is also average, that risk is… 60 percent (in the American population). To sum up: everything is relative. But it is also clear that my risks are also affected by my life-style, my diet, etc. – and the good side is that I have ways at my disposal to minimize those risks.


Brown eyes

Concerning my physical traits, I learned that I “probably have brown eyes”. In fact, they’re green. My father had brown eyes, but not me, I’m sure of that. Nevertheless, green appears, in my report, as third choice for eye color, following brown and blue. When I saw the result, I said to myself: “If they don’t get it right for eye color, should I trust the rest?” But the fact is, as they explain on that same page, that in spite of being almost totally inherited, eye color is controlled by a whole bunch of genes. That means that the genetics of eye color is very complex and most of the genes involved are still unknown. Another surprise was discovering that my genes dictate that it is likely that I don’t have many freckles or moles. I couldn’t help laughing – I have a lot of freckles and moles. But the fact is that skin complexion and eye color are related, so it’s not surprising they also got that wrong.

Anyway, my genes and I agree on other things: they tell me I’m lactose-intolerant (and it’s true that drinking milk makes me nauseous); that I’m susceptible to viral gastroenteritis (as I know by experience); that my earwax is wet (right!); that I don’t flush when I drink alcohol (right!). On the other hand, they also inform me that I shouldn’t like the bitter taste of broccoli or of unsugared coffee – but I do; that I’m not immune to HIV or malaria (I hope never to have the opportunity to confirm this); that I have sprinter’s muscles (maybe I should take up jogging… but I probably won’t). One last result – and here I trust genetics: I don’t carry the mutation for cystic fibrosis, a very serious genetic disease. This information is important, because carriers, even if they don’t have the disease themselves, can transmit it to their offspring.

There is still room, in this long list, for a series of cognitive attributes, such as “measures of intelligence” (it seems I’m smart!); memory (it seems I have a good memory!). They also tell me I am capable of learning from experience. But here, we are obviously treading a very unsteady terrain.


Genetic roots

As I already explained, my main motivation, when I ordered the kit, was to find out something about my ancestors – or rather, about my maternal descent. Being a woman, I didn’t inherit my father’s Y chromosome and so my DNA cannot give me unequivocal information about my paternal descent. To achieve that, I would have to have my brother or one of his sons test their own DNA – or at least their Y – but that’s another story.

In order to determine my "matrilineal inheritance”, at 23andme they analized my “mitochondrial DNA”, a part of my genetic makeup that isn’t found inside cell nuclei like the rest, but rather inside cellular organelles called mitochondria, which are the cells’ batteries. These are transmitted, unaltered, by mothers to their sons and daughters, inside the egg’s “white”. Contrary to what happens with the rest of the genome – where half the information comes from the father through the spermatozoid and the other half from the mother through the egg "yolk" – the mitochondria and their set of genes come exclusively from the mother.

Mitochondrial DNA, which contains around 16,000 base pairs, was given to our mother by her mother, and to her mother by her grand-mother, and so on. As time and generations pass, that little bit of DNA mutates with a frequency that is more or less known. That’s why it's been used by geneticists as a “molecular clock”, to go back in time (and space) to the so-called “mitochondrial Eve”, the mother of all modern human beings. Thanks to the study of mitochondrial DNA, it is today widely accepted that that primordial woman lived in Africa some 200,000 years ago.

Every time a new mutation arises in a woman’s mitochondrial DNA, and that “founding mother” passes it on to her children, that gives rise, in theory, to a new motherline. But not all those lines survive: only some of those genetic families – or “haplogroups”, as experts call them – survived to the present day in the genes of the different peoples of the world. And by comparing the different mitochondrial DNA lineages which exist in current populations, it is possible for scientists to go back in time and space to approximately determine the temporal and geographical origin of different haplogroups and reconstruct the path of human migrations through the millennia.

At 23andme, they read some 3,000 SNPs distributed along the whole sequence of the mitochondrial DNA to determine a person’s haplogroup – his or her family by matrilineal descent. Sometimes, the haplogroup assignment can be combined with historical data to provide information about the possible whereabouts of someone’s ancestors a mere one or two thousand years ago.


Genealogical disappointment

However, as it turned out, my haplogroup, called H7, seems to be very rare (which simply means, I suppose, that not enough people belonging to that haplogroup have had their mitochondrial DNA tested yet). Because of this, the geographical origin and the age of my haplogroup are still pretty vague.

I must confess I felt pretty disappointed. I had imagined that I would find out how and where and when that early founding mother of which I am a direct descendant might have lived. But it wasn’t to be so. Pity.

Meanwhile, during the last few weeks, I have consulted a series of online community discussions at 23andme, I’ve performed Google searches, I’ve written emails to several people. I discovered that there were ways to compare my mitochondrial genes to those of others, but I haven’t been able to actually do that yet. I joined a Yahoo! mailing list of people who are also H7, but some of the information I got through that channel seems to contradict other informations I gleaned from other websites.

What I know, for the time being, is that haplogroup H7 is an offshoot of another haplogroup, H, which is today the most common haplogroup in Europe and which, according to 23andme, originated in the Middle East around 35,000 years ago. Then, 25,000 years ago, it spread across Europe. Several thousand years later, at the peak of the last glaciation, or Last Glacial Maximum, those men and women were forced to take refuge in the more temperate regions of the continent – the Iberian Peninsula, Italy, the Caucasus. What happened next is suggested by several studies – one of them by António Amorim and his team at IPATIMUP (Institute of Molecular Pathology and Immunology), at the University of Porto, Portugal: when the ice started regressing, about 15,000 years ago, the members of haplogroup H may have started to reconquer Europe, giving rise, after several later tribulations, to different sub-haplogroups, among which H7.


A soupful of letters?

Another thing I found out about my genes’ “geography” (and not only the mitochondrial set) is that, although more that 99 percent of my chromosomes denote a European origin, a small portion, less than 1 percent, comes from Asia. I'm really curious about the specific geographical origin of that tiny bit of DNA from another continent…

I was also able to confirm, thanks to a very interesting functionality at the 23andme website – the so-called “global similarity” announced in the welcome page – that on a global level, the population my chromosomes seem to be closest to is centered on the Ukraine. This doesn’t surprise me, as it is precisely from the Ukraine that both sides of my family, both maternal and paternal, come.

However, one could object that, overall, I don’t know now much more than I already knew. After all, just by looking at the life story of my parents, grand-parents and so on, I manage to get a pretty good idea of the inheritable pathologies that might threaten me in the course of my own life; I don’t need a genetic profile for that. And neither do I need to know my genes to know that it’s better to lead a healthy, active life and all that. Regarding my origins, they remain as obscure – or as clear – as they were before. Ultimately, the data I obtained is merely a gigantic soupful of letters.

But that’s not exactly the way things are; this state of affairs is not fixed. As new, credible information comes along about other SNPs among the 550 million that were analized, 23andme will quickly update my profile to include them. The service we pay for includes updates of the interpretation of our data in the light of the most recent scientific results. What I can ask, instead, is what I’ll do when they offer to let their clients know, as they’ve already announced they would, about their risk for Alzheimer’s disease. I still don’t know what I’ll do. What I do know is that, in spite of the obvious limitations of this new science, I’m still as fascinated by what I imagine my genetic data might tell me in the future. And that’s like embarking on a new adventure each day.