December 19, 2013

How is non-invasive prenatal testing getting on?

Almost two years ago, I wrote a series of posts on non-invasive prenatal testing (NIPT) of foetuses and the associated market. What has changed since?

In terms of technology, there hasn't been any big breakthrough. The principle behind NIPT has stayed the same: Some of the DNA in a pregnant woman's blood is of foetal origin. By sequencing that DNA, it is possible to reconstruct most of the genotype of the foetus, and to detect many severe genetic disorders from the 8th week of pregnancy onwards - earlier than with equivalent tests.

What has changed is the number of companies actually offering NIPT. At least seven companies worldwide currently sell NIPT, whilst several other are aiming for a 2014 market entry. The table below provides and overview.

Company (Test)
Conditions tested
Test price
Tests sold
Market entry
Ariosa (Harmony)
Trisomies 13, 18, 21, sex chromosome aneuploidies
150,000 until September 2013
May 2012
Beijing Berry Genomics (BambniTest)
Trisomies 13, 18, 21
Around $500

China, Worldwide
Trisomies 13, 18, 21

200,000 until October 2013

CellScape (Clarity)
In development

None yet

In development

None yet

LifeCodexx/ GATC (PraenaTest)
Trisomies 13, 16, 21
€1,150 ($1,580)
Year to August 2013: 6,000
August 2012
Natera (Panorama)
Trisomies 13, 18, 21, sex chromosome aneuploidies

December 2012
Sequenom (MaterniT21 Plus)
Trisomies 16, 21, 22, deletions causing various syndromes
$1,700 for uninsured patients
145,000 until September 2013
October 2011
Verinata/ Illumina (Verifi)
Trisomies 13, 18, 21, sex chromosome aneuploidies

March 2012

This list is incomplete in the sense that increasingly NIPT is also considered as an option by public health services. In the Netherlands, a NIPT trial will start in April 2014. In England, the National Health Service (NHS) is currently carrying out a study of NIPT called RAPID.

One aim of the NHS study is to compare the cost of NIPT to that of chorionic villus sampling (CVS). CVS is an invasive procedure that has traditionally been used for obtaining genetic material from the foetus. It involves taking a sample of the placenta and can lead to miscarriage. The advantage of CVS is that it gives relatively unambiguous results, whilst positive NIPT results still need to be confirmed by CVS or a similar procedure. Even so, the estimated total cost of using NIPT compared to CVS alone was 4% lower for one condition and 7% lower for the other condition. An important caveat of this estimate is that it only used NIPT for foetal sex determination in cases where this was disease relevant and not into detecting chromosomal abnormalities. 

The Dutch and English examples of public health service-provided NIPT points to a mistake that is easy to make when estimating the future market size for NIPT (and genetic diagnosis in general). A lot of health service providers may prefer to develop their own tests rather than rely on third party tests. This decreases the size of the market accessible to companies like Sequenom and Ariosa, especially in Europe.

Do you have any comments, or are there any NIPT companies or projects I haven't included? Please feel free to tell me in the comments section below.

November 25, 2013

Should 23andMe be regulated?

Today, the American Food and Drugs Administration (FDA) has sent a letter to 23andMe, the market leader in direct-to-consumer testing, asking them to immediately discontinue marketing their tests. Is it justified in doing so?

In my opinion, the answer is clearly yes.

That's although 23andMe has, I think, not marketed its services in any unethical way. Their website and their most recent TV ad avoid raising fears. This is in welcome contrast to many adverts by established pharmaceutical companies I see in the United States.

My point is that 23andMe's test should not be treated any differently than other diagnostic tests.

But is 23andMe a diagnostic test? On its website, the company lists the traits and diseases it returns information on. Whilst some of them, like ancestry or odour sensitivity, are harmless, others are specifically designed to inform on the risk for diseases including Alzheimer's, breast cancer, Parkinson's, and many more.

To claim that all 23andMe does is return information about variant-gene associations is disingenuous. By extension, there shouldn't be regulation of any diagnostic test, because all they do is return information on some marker-disease association. This may be a valid point of view, and if you hold it, you need to argue for the deregulation of the diagnostics market as a whole, and not just against the regulation of 23andMe.

A few weeks ago, I bought a test from 23andMe myself, but I haven't received the results yet. Does this make me a hypocrite? I don't think so.
Firstly, I strongly advocate that people have free access to their genetic information, and to their medical data in general. That does not mean that the quality of disease-relevant genetic tests should not have to meet the same requirements than other diagnostic tests.
Secondly, I feel that because I am working in the field, I have a good grasp of how reliable genotyping data as that provided by 23andMe is and what it can tell me. If my test returned anything that was of serious concern to me, I wouldn't dream of taking drastic action without further testing to confirm the results. I believe that this distinguishes me from some other people, including the least educated and vulnerable ones.

Some people may feel that the FDA has an anti-genomics agenda. I don't know whether that's true, but I don't see the letter it sent to 23andMe as any evidence for it. If genetic testing that fulfils regulatory requirements for $99 is not possible, that's unfortunate, but why should an exception be made in this case?

The opinions I express on this blog are not necessarily those of my employers or my funding body.

October 19, 2013

What's the gene for music appreciation?

This is a post on why there is a lack of good explanations for the things that make life worth living.

Why do we like the things we like? For some, such as food or sex, the answer is  obvious. Without them, we wouldn't survive or reproduce. In order to make sure we comply with its wishes, our body uses a primitive carrot and stick management approach consisting of rewards (e.g. sweet taste) and punishments (e.g. hunger). Why our body can't treat us like adults I don't know. Maybe it's so that we don't end up like people who have a mutation that prevents them from feeling pain and therefore are prone to self-inflicted injuries.

There are however plenty of things we like that can't easily be explained this way. For example: Music, truth, humour, stories and drama, beauty, and spirituality. For none of these it is obvious how liking them promotes survival or fitness.

Let's consider music. Nobody I've talked to about this has been able to satisfactorily answer the question of why we like it. The question is not just why we like it, but also why we don't like other noise that is very much like music. There are certain tones and combination of tones we find harmonious and like to listen to, whilst others we just find annoying. This preference is somewhat related to things like frequency. For example, tones that have a multiple of the frequency of others appear to be harmonious, and why that should be so is not clear.

There is of course a cultural element to music, but there also seem to be universals. People are divided on whether they like techno, but I assume nobody finds waltzes unpleasant. Another observation is that music is somehow connected to dance and language, but what exactly these connections are and whether they are essential is unclear as well.

One possibility is that music appreciation is a by-product of some other capacity that our ancestors evolved, such as language. This explanation is advanced by some people working in the field of neuroaesthetics. To me it seems unlikely that our brain is so inflexible that it couldn't evolve one capacity without also accidentally evolving another.

I have also come across the theory that music appreciation is nothing but pattern recognition. To me, this is not satisfactory either, as it does not explain why we find music more pleasant than other instances of pattern recognition. For example, contemplating visual patterns can be nice, but I don't know anyone who spend hours every day doing it for recreational purposes.

Evolutionary psychology could explain our like of music and similar phenomena through sexual selection. According to this theory, music is like a peacock's tail: Potential partners of the opposite sex like a nice tail (or good musical sense), which means that having that tail (or sense) increases our chances of reproduction. Over time, the tails (and musical sense) get more and more elaborate. Whilst this is a valid explanation, and I admit that it seems to apply to music, it's not satisfying either. Why? Because absolutely every trait could be explained this way.

If this were a well-considered blog post, this would be the point where I advance a hypothesis that may provide an answer, or at least a proposal how an answer could be found. Unfortunately, it isn't and I can't. All I can do is to lament the fact that there is a lack of good explanations and leave it at that.

October 6, 2013

Is direct to consumer genetic testing profitable?

Direct to consumer DNA testing is attracting a lot of attention. Is it ethical, how should it be regulated, does it make medical sense? These are all interesting questions, but the one I'll try to answer in this post is whether genome testing makes business sense.

The most well-known company in this space is 23andMe. That's no coincidence: It is working hard at increasing its profile and even has its own commercials.

As of September this year, 23andMe has genotyped 400,000 customers. That's 220,000 more than a year earlier, meaning that at a current price of $99 per test 23andMe made around $22m in the last 12 months. Contrasting this with the $161m in investment the company has attracted since it was founded in 2007, that's not great.

Competing companies are not doing much better. As the list below shows, 23andMe is currently to my knowledge the only company still active selling direct-to-consumer genetic testing services. The others have either declared bankruptcy or have otherwise stopped offering direct-to-consumer services.
It's obvious that genetic testing by genotyping is a niche market. Since it's unlikely that the value of these tests to consumers is going to increase dramatically, it's not clear to me how this could change in the near future.

September 9, 2013

Has the cost of sequencing stabilised?

Almost two years ago, I asked how DNA sequencing costs would develop in the future. Extrapolating from the data that was available at the time, my prediction was that by now, the cost of sequencing a human genome would be below $100.

As it turns out, the future is not an extrapolation of the past. Since early 2012, the cost of sequencing has remained stable. This is more remarkable than it seems at first, because before 2012, the cost of sequencing had declined significantly in every year since 2001. But since the beginning of 2012, it hasn't budged and still remains at $6,000 per genome.

What went wrong? The biggest thing that happened was nothing. There is just not enough innovation in the sequencing market. Even if the two companies that dominate the market, Illumina and Life Technologies, came up with a vastly cheaper technology, it would probably not make commercial sense for them to make it available right now.

Eventually, there'll be some new technology that will trigger a new round of innovation, but right now it is not clear from where this will come. Nanopore sequencing still seems like a possible candidate, but that has been the case for a long time.

May 15, 2013

Whatever happened to genetic matchmaking?

Online dating websites are very lucrative, and the competition between them is immense. Every other advert on the London Underground seems to be for a dating website catering to one demographic or the other.

A few years ago, a number of startups began to offer online dating based on genetic profiles. Members would send in a sample of their saliva, from which DNA would be extracted. Next, the genes encoding the Major Histocompatibility Complex (MHC) would be genotyped and added to the member's profile. The theory was that we are sexually attracted to partners who have different MHC genes to ours. An algorithm would then identify partners that were a good genetic match.

There were at least half a dozen startups that tried this, including, Gmatch, Basisnote, and GenePartner. All of them, with the exception of GenePartner, are now out of business.

Why didn't this business model work? I don't know about you, but for me, it's a lot more fun to think about why a business might fail instead of why it might succeed. I have several hypotheses:
  • It's not particularly romantic to find your partner on the internet. Finding your partner based on a genetic test that is likely to be obscure to most people is even less romantic. Actually, it'd be hard for me to come up with a less romantic scenario.
  • Most people (including me) will assume that things like personality, interests and looks are more important in a partner than their MHC genotype. 
  • Presumably dating websites have to balance their male and female membership. My guess is that the majority of people interested in finding a partner through genetic matchmaking are male. 
  • This type of service may attract people with certain undatable characteristics. 
  • The science behind genetic matchmaking is shaky. I only reluctantly include this point because I consider this an unlikely reason for business failure. After all, shaky science doesn't stop other online dating websites that use unvalidated algorithms.
Unfortunately, I have no privileged insight into why the genetic matchmaking startups actually failed, and I'd therefore love to hear from you if you have any other ideas. 

Disclaimer: My interest in genetic matchmaking is entirely professional in nature. However, should you avail yourself to any of the still functional services I have mentioned in this post, I'm not going to judge you. At least I will try not to.

April 19, 2013

What's wrong with writing a business plan for investors?

I've recently been involved in writing a business plan. Whilst it was a worthwhile experience, it has also prompted me to question whether writing a business plan for investors - rather than for internal use - is in anyone's interest.

We were writing for a biotech business plan competition. The winning team was to get £100,000 ($150,000) and free lab or office space in a biotech incubator. The judges of the competition consisted of venture capitalists (VCs) and other investors, as well as other organizations that are involved in early-stage startups. The selection process was quite similar to that used by VCs when they decide which businesses to invest in.

In principle, there are sound reasons for writing a business plan. It forces the startup to envisage its future and to think through the details of what it will do. Ideally, this process will uncover potential threats and opportunities, strengths and weaknesses.

This is however not what happens in reality. In my experience, people starting a business already tend to be overoptimistic. Unfortunately, I doubt that writing a business plan for potential investors encourages honest and self-critical assessment of how to maximise the startup's chances of success. 

Instead, it becomes an exercise in trying to convince potential investors. When there are several estimates of the market size, the temptation is to pick the biggest one. When there are potential non-obvious threats, the temptation is to either not mention them at all or to explain why they are not going to be a problem, rather than actually thinking them through. Et cetera. Some investors even encourage this by explicitly demanding to be "sold" the idea.

You may think that this is not a big deal. After all, investors are aware that startups will inflate their chances of success and will discount for this: You're hyping, I know that you're hyping, and you know that I know that you're hyping.

The problem with this is that apart from money, time is the most limited resource for startups. Writing a business plan takes time, and the one the startup writes for investors is probably going to be the only one. Writing a second, more realistic one for internal use would be a waste of resources.

A solution would be for investors to stop asking for a business plans and instead relying exclusively on their own due diligence - in close collaboration with the startup of course. The initial communication between the startup and the investor would focus on the science, covering the market and the business model only in general terms. The advantage for startups would be that they could write their business plan without having to keep in mind what the investors want to hear, instead using their resources on planning that actually matters to the business.

If you have any experience of investing in early-stage startups, or if you have ever written a business plan yourself, I'd love to hear what you think about this.

April 9, 2013

What would free sequencing mean?

This is an essay I wrote for a science writing competition before realising that I'm not eligible to enter. Please don't mock me.

Do you remember the first time you used a computer? I do. It was around 1990, and I was six years old. My father had taken me to his office, and one of the secretaries showed me her PC. I don't remember much about it except that several times whilst she was working on it it gave off a tortured noise to indicate that it needed more time to catch up with her typing.

Today any smartphone has hundreds of times more computing power at a fraction of the cost. That computers steadily get cheaper and more powerful has been recognised for some time now and the phenomenon even has its own name: Moore's law.

Moore's law, interesting as it is, pales in comparison to what is going on the world of DNA sequencing.

The Human Genome Project, which had the goal of sequencing the DNA of a single individual, took more than ten years and cost roughly $3 billion. Today, it costs a few thousand dollars to sequence a genome, and it can be done in a few hours. Over the last years, the pace at which DNA sequencing technology had become cheaper and better has been staggering, and there is no indication of it slowing down. On the contrary, there are several companies that work on innovations that will ensure that sequencing will become faster, and easier to use, and much, much cheaper. The most likely scenario is that in a dozen years or so, sequencing a human genome will essentially be free.

It is worth contemplating for a moment what the implications of that will be. Depending on your attitude towards technological progress in general and genetics in particular, the following may either excite you or make your very anxious.

Sequencing will initially have its biggest impact in the area of healthcare. Already, sequencing enables the accurate diagnosis of children with severe genetic disorders and of adults with certain types of cancers, and in some cases has led to cures that would not have been possible otherwise. 

Another example of how sequencing is already impacting healthcare is the diagnosis of genetic disease in unborn babies as early as in the 10th week of pregnancy. Through the widespread adoption of this kind of tests a lot of severe genetic disorders could be avoided, although this is also likely to lead to an increase in the number of abortions. As time goes by, it'll be likely that sequencing becomes an increasingly important part of everyday healthcare.

Cheap and easy to use sequencing technology may however have its biggest impact outside of the hospital. Computers only really started to change the world once they got cheap enough to become affordable toys for teenage nerds and hobbyists. It may well be that the same is going to happen with DNA sequencers. Handheld sequencing devices could be used by hobbyist entomologists who want to identify bugs in their back yard and by consumers who want to make absolutely sure that their lasagne contains nothing but beef.

The impact of pervasive sequencing could also have more sinister aspects. Today, a lot of effort goes into protecting the data that is obtained by sequencing clinical patients from being accessed by unauthorised third parties. Imagine anyone with a handheld sequencer could get the same data about you from some dandruff you have left behind on your keyboard.

January 26, 2013

Why has Seqonomics been so quiet recently?

If you come here often, you'll have noticed that I recently haven't kept up with posting every week. The reason is that I'm currently working on another project in my spare time.

In the end of March, I'm going to launch Consultinsider, a resource to help graduates, postgraduates and postdocs start a career in management consulting.

The provisional Consultinsider website

If you like the idea, I'd appreciate your help. Here is what you can do to get the word out: 

But don't worry, Seqonomics isn't going away. Once Consultinsider is up and running, posts will commence once again.

January 11, 2013

What is going on with the UK?

The United Kingdom is punching above its weight in genomics.

Compared to the other two European countries I've covered on this blog - France and Germany - the UK is the clear genomics leader in genomics. The reason is not so much that the rest of Europe has neglected genomics, but that the UK has gone all in.

The Sanger Institute, where I work, is a good example of this. It made the single largest contribution to the Human Genome Project, and it is still going strong. Together with the Broad Institute, it is today the leading genomics institute worldwide in terms of high-impact citations, and amongst the largest sequencing centres in the world. It is largely funded not by the British government, but by the Wellcome Trust, a charitable foundation.

Sequencing technology is another example. Both the most prominent sequencing technology used today and the technology that many see as most promising for the future were developed in the UK. Sequencing by synthesis, which produces most sequencing data nowadays, was developed at the University of Cambridge before being spun out into a company called Solexa, which was acquired by Illumina in 2007. The company that is most likely to first bring nanopore sequencing to the market, Oxford Nanopore, is also located in the UK.

The British government is keen to keep and expand the UK's capabilities in genomics. Last December, it announced plans to sequence up to 100,000 patients in the NHS, the UK's publicly funded healthcare system. It has not announced many details, except that it has earmarked £100 million for this project. This money will not only cover the sequencing, but also training and the development of bioinformatics capacity. This is about all the information that has been made public about this project, but a more detailed plan is expected by June this year.

Especially compared to similarly sized countries such as France, Germany, or Japan, the UK has a strong base in genomics, and there seems to be enough public commitment to ensure that this is not going to change over the next few years.

This post is a part of a series on genomics in different countries, which has already covered Canada, China, France, Germany, and Japan.