market research

[smice]

Hello everyone!

I need some help. I have to introduce the growing market of bioinformatics (and ng sequencing data analysis, specifically) in a few pages, but I cannot really find any useful reports on the market, nor up to date figures. What I've found are some comprehensive books, which are surely great, but I cannot spend thousands of dollars to buy them and weeks of reading to wade through thousands of pages. I've also seen some rough (and sometimes contradictory) estimations about the market in 2004 (and before), but no exact figures about the current market and the future prospects.

Can anyone help me please?

I am interested in any kind of usable data, including the description of the bioinformatics market as a whole, but especially in a detailed analysis about the market of (next generation sequencer) data analyzer software tools.

Thank you and best regards,

smice

[apfejes]

Hi Smice,

From my time in industry, I recall working on similar projects. Unfortunately, the vast majority of the information is collected by people who are looking to sell it - and the numbers are difficult to come by otherwise.

Additionally, your question is pretty vague. Bioinformatics is really a tool, not a market: It's roughly equivalent to asking what the market is for statistics or chemistry. You'll probably have to refine your question significantly before you'll be able to come up with any form of reasonable analysis.

As for looking for Next Gen sequencing markets, I suspect it's just changing too rapidly for anyone to give you a number that will make sense. I expect that the only people who have a good guess as to what that value might be (other than the organizations that collect this for resale) are the companies who are already in that space, and thus are probably not willing to share their market data.

My advice would be to see if you can find the SEC filings for any of the companies who are in this business, and see if you can tease out the trend from their revenues. Since most of the platforms have been acquired by larger pharma/sequencing groups, it may be hard to isolate the numbers you're looking for, anyhow.

Good luck with your search.

[smice]

Thank you for your answer and the advices!

By the way, I know the question is vague, but the analysis of the bioinformatics market is not MY question. I would be most happy if I could focus on a narrower field.

[Joann]

signposts

If I had to take a shot at researching this question I would first keep in mind that
1. as opposed to traditional sequencing methods in biology, next gen sequencing is entirely bioinformatics driven. Furthermore,
2. I would expect that any lab or center already engaged in traditional sequencing will want to move forward with next gen methods.
3. I would take stock of what's happening with open access repositories for next gen sequencing data.
4. Assume that any and every DTC (direct to consumer) gene sequencing operation will be working with next gen sequencing, sooner or later.
5. Spot major sequencing centers (see other forums in this site)
6. Do some sort of survey with published literatures.

Hang some numbers on the above framework.
Hope this makes it more interesting for you.

[apfejes]

Hi Smice,

I'm sure I'm preaching to the converted, but what the heck - this is a topic I enjoy.

It's aways the same thing: the way market questions are brought forward by the investors/start-up companies is always vague. I think the underlying question really should be "what is the right market to be positioning ourselves in?", and thus, part of this research is to guide the people who are asking to the right set of questions.

The best way to approach this is generally to ask what the opportunities are, and then what the value of each of those markets will be. For instance, within bioinformatics, there are only certain needs: Lims systems, equipment support, next-gen sequencing support (aligners) and specialty applications (molecular modeling). (I'm probably forgetting a few.)

My research model would differ from the traditional one that Joann has suggested: What she's suggesting is to figure out the maximum market potential for Next Generation Sequencing Hardware, which is only a relevant number if you will be producing a next-gen sequencing platform of your own.

If you follow Joann's method, you'll end up with a vast over-estimate of the money that will be spent on next-gen hardware/software. Not all sequencing centres will drop traditional sequencing: there are still applications for which sanger is the right solution. Not all DTC centers will switch to next gen - arrays will still provide cheaper solutions for identifying specific small sets of SNPs for at least another 4-5 years. Finally, the vast majority of bioinformatics needs of the sequencing centres will probably be supported by academic software and in-house development. (And, of course, the business model for the sequencing centres that do exist is rapidly changing, so it's pretty much impossible to predict what the need for the machines actually will be.)

I think it's fairly clear that the companies who jumped in to the bioinformatics field thinking that they could dominate the market with a fast aligner are not making money at the rate they had hoped for. (eg, compare the rate of adoption of ZOOM, or NovoAlign compared to MAQ/BWA.) I suspect many of them had done the math and come up with much larger market numbers that they were able to realize.

Anyhow, which ever way you chose, this is certainly an interesting topic, and the answers that anyone comes up with involve a lot of forecasting and modeling. No matter how you get there, it still involves a best guess and many leaps of faith.

Best of luck,

Anthony

[Joann]

counter point

From my background as a basic biologist, I would like to counter some of Anthony's points because the applications of computer technologies to biological problems so far have produced some absolutely unimaginable advances. For higher organisms, unlike bacteria, it's not just the genes that are going to be relevant. Whatever else may come to light from our dissections via massive sequencing, there is a good chance that sequencing will go on to serve as the plain vanilla viewing microscope. Add to this the growing commercially relevant work (in terms of biotech patents, for example) on genes, regulatory regions, and specialized organisms that really can't be contained by academic centers. All of this is supported by bioinfo expertise.

[apfejes]

Hi Joann,

I think your underlying assumption is that all sequencing is going to be replaced by next generation sequencing is somewhat overstating the case. I've always considered Illumina/454/SOLiD sequencing as the "second generation" of sequencing technologies. They do reasonably short reads with reasonable accuracy using one or another form of pyrosequencing. That's great, however, they really are not applicable to all forms of genetic research.

PCR isn't going to disappear completely because of these technologies, and the "third generation" of sequencing technologies is already following closely behind. (See Pacific Bioscience's SMRT technology.) We're in the middle of a transition period where each of these three generations will be competing to find their best niche, along with other technologies.

PCR will always rule the domain of cloning and construction, Arrays will always rule the diagnostic field, second generation sequencing will probably remain dominant in ChIP-Seq/RNA-Seq, while whole genome resequencing will probably fall to the third generation platforms. If you collect the numbers as you suggested in your first post, you'll be neglecting the incredible diversity of the field that exists and that will completely miss the trends of changing technologies.

Sequencing *will* continue to underly the basic research, but lets not assume that the markets for the technology will all jump the same way. None of the players in the field has assumed dominance - and none likely will in the short term.

Nor, for that matter, will bioinformatics be a field in which there are huge profits to be made. In the software business, the barrier to entry is far too low for your competition unless you have either a head start or some secret recipe that no one can copy. (Or are willing to engage in unethical practices like Microsoft.) In the world of next-gen sequencing, I think all of those are unlikely.

Perhaps we should agree to disagree: Sequencing in all its forms will continue to power the research platforms of the future, but the next-gen after-market for bioinformatics products will remain a desert for companies looking for profits.

[Joann]

Anthony, I think your points are as good as mine as far as future business prospects may be ascertained. What I wanted to emphasize is that as far as unsolved biological problems, I mean sort of like the questions that haven't even been asked yet, biologists seem to really need to go to the powers of next gen sequencing, (and the totally computer supported sequencing platforms beyond that) methods for either large chunks or a whole genome. Because computers can right now put back together all the pieces from a whole genome, discovery steps in between grabbing one gene and all of the chromosomes whole have been passed over--just because of the nature of the computing technology and what it offers. The fact is there is surprisingly little published work on physical genome fractionation methodologies--I think that's because biologists are mostly thinking, heck just sequence the whole thing. There is elegant methodology in all of the niche applications you described and they will remain useful for answering the biological questions generated out of each technology, but every species and species variation has a genome, not just genes, preserved over eons of evolution and there is no good answer to what's it for? yet. The obvious approach is to do lots and lots of sequencing. And that's a lot of computing. And a lot of work. Maybe not a whole lot of profit, unless a company works very hard. So what's new?

[apfejes]

Hi Joann,

I suppose we both agree, then: There's lots of sequencing to be done.

My guess is that it won't be the companies doing most of the sequencing, but rather the academically supported labs, since there's very little profit to be made sequencing things - unless you can use that information to support a drug passing through the FDA. (One of the few sequencing applications that is likely to have any serious level of investment in the future.)

Like most other science fields, the basic research (eg, sequencing or bioinformatics) isn't what pays the dividends - it will be the application of the knowledge learned from them that creates products that bring in profit (eg, new drugs, better health products, etc).

Just to bring this full circle, bioinformatics as a market is practically non-existent, however, as you've pointed out, Bioinformatics as a research tool is incredibly powerful. As long as we can keep the two concepts separate, then clearly we're all on the same page.

Cheers,
Anthony

[Joann]

more than that

Anthony,
You're just not giving enough credit on the computer side of bioinfo and I'm not gonna let you get away with that! Health is just one area of biotechnology where serious bioinformatics supported progress can be expected. Genes and genetics are the ultimate source of all natural products under the sun, like the species and gene expression patterns that produce great-tasting coffee, for example. What's not commercial there? I'm just explaining that bioinformatics-supported sequencing is shaping up on the biological understanding side to be an unprecedented key tool of inquiry and that biology can no longer be done without bioinformatics. Perhaps bioinformatic sequencers should take a cue from the natural world and start aggregating themselves into more compelex working units!

[apfejes]

Hey Joann,

I'm really enjoying this conversation! (=

I happen to believe that you're right, that genes and genetics will underpin an entire economy of products... however, the bioinformatics that gets us there (for the most part) won't be available as an off the shelf commodity.

As an example, if you look around for companies making molecular modeling software, very few of them are doing really well: There isn't a thriving market for molecular modeling tools. People are eeking out a living with docking or whatever, but the market isn't in building bioinformatics applications. In contrast, there are companies doing well using these tools to solve problems: for them, the bioinformatics is in-house, as a side product of their biology goals.

The example I have handy is Zymeworks (Zymeworks.com), where molecular modeling is used to develop structurally modified proteins. If they had tried to sell a molecular modeling platform, I doubt they'd have gotten any traction. Instead, they built a tool that allows them to address biology questions. Their edge may be in bioinformatics, but it's not a bioinformatics product that will earn them any money. (Disclosure: I hold shares in Zymeworks.)

I also agree that bioinformatics is an unprecedented tool for inquiry and understanding, but I disagree that there is money in it in the next 10 years. While Second and Third Generation sequencing may create an appetite for bioinformatics, it will not create a viable market for it until someone finds a way to patent pure algorithms, which I hope never happens!

Each time a commercial product exists for a particular application, it won't be long until an academic or open source product matches it's functionality and becomes available to displace it. Bioinformatics isn't like word processing - you can't sell millions of copies of a bioinformtics application: which means you need to sell it at a high profit margin to a small number of people. Putting a large margin on it means that (for most academic labs) it will inevitably be cheaper to hire a couple of grad students to re-write the algorithm in-house, which inevitably means more competition, which means an eroding of the margins. There just isn't a way around that.

Anthony

[Joann]

some additional points

Well Anthony, biologists like to specialize and I believe there is lots of room for specialized bioinfo operations to develop. As commercial avenues evolve, chances are there won't be an available grad student with an interest in every particular specialized project/product. On the corporate side, right now lots of drug companies are in no-man's land. The chemical pipeline seems to be drying up but kind of like GM nobody wants to make that paradigm shift. Bioinfo leaders could help transition major companies in a major way, but it could be necessary to collaborate initially with the past way of doing things to smoothe the pathway. Biologicals are patentable and there is no defined regulations yet for dealing with generic biologicals. And biology research and development is often ongoing without ever solving the problem. Since wet lab biology research is always very expensive, institutional infrastructure is going to be necessary for that side of the equation.

[apfejes]

Hi Joann,

I disagree with you on two related but substantial points.

The first is a logical contradiction. While you claim that the biologicals are valuable (which I also believe), and that there will be bioinformatics tools that will help us get at those biologicals faster or more efficiently (which I also agree with), I'm not sure how you can then infer that the bioinformatics is also valuable to the market. (Building a very expensive piece of art out of a piece of stone may give rise to a very valuable end product, but doesn't mean that all stones have a great value.)

Anyhow, my second objection (which is closely linked to the first) is that I don't believe there are any sustainable business models for companies that would like to profit on bioinformatics alone. After spending a few years looking, the only business models I could make work were those that involved not selling bioinformatics tools. If you can come up with one, and are willing to share it (in public or private), I would concede the argument. (=

The only protections available to software are trademark and copyright, neither of which can prevent competition. How can a bioinformatics company make sustainable profit when they are forced to sell their software to a small market at high cost with no barriers to entry for their competition?

Just to rebut another point in passing, I don't consider LIMS systems (about the only profitable next-gen related institutional infrastructure that sells well of which I'm aware) to be a bioinformatics application - so I'm not going to accept that as an example of a bioinformatics business model at work. (-:

Cheers,
Anthony

[bioinfosm]

Not sure which side of the argument it supports, but the point according to me for a bioinformatics biz model is to sell the service. There are no golden rules or tools for various projects. Even with sequencing, I see so much variation in different projects, that one has to do things differently, even with the same tools.

So essentially, a charge-out model to charge for expert bioinformatics services might be one way to look at it. A by-product is new tools or workflows, that can be marketed or used in-house for increased efficiency.

btw, it is nice to follow this discussion

[apfejes]

Hey Bioinfosm,

Thanks for contributing - I think you raise a good point. Bioinformatics support will probably be a strong market, although I strongly suspect that most companies will need to be constantly updating their pipelines/bioinformatics processes, which makes it far more attractive to bring that type of support in-house in the long run.

Still, for specialized applications, there may be some value there.

Unfortunately, then only person I knew who was doing bioinformatics consulting was recently picked up by a larger company so I'm still not convinced how sustainable that model is in the long run.