Thursday, June 27, 2013

Price War Between Sequencers Continues

Elilene Zimmerman at CNNMoney writes:
Entrepreneurs and scientists are pursuing an even more dramatic medical breakthrough: The ability to sequence an entire human genome for around that same $100 price tag. That goal remains a few years away, but the obstacles are falling fast.

Sequencing is a way of "reading" DNA molecules -- two strands twisted together to form that famous double helix. The entire human genome contains roughly 3 billion molecular base pairs, which researchers study to find variations that might play a role in the development of diseases. Right now it typically costs $1,000 to $4,000 to map out an individual's genome. (Specialized sequencing -- for, say, a cancer patient -- often costs more.) 
A lot of great things can be achieved by embracing a "cheaper, better, faster" approach but it seems that too much attention is paid to the cost per genome sequenced.  The price tag is a neatly packaged quantitative measure that's easily quotable, but it doesn't convey the whole story.

What people actually using the "$1000 genome" data will tell you, from experience, is that cheaper genomes are not necessarily better. 

For instance, a lot of variables can be tweaked to reduce "price per genome" to yield market-beating prices but delivering sub-standard data quality (I don't mean that it's poor quality data, just data that doesn't meet the standards of what's current in research at the time).  For instance:
Right now [Ion Torrent] can sequence the exome -- the 1 percent of the genome we know how to interpret -- for $500. "In three months, we'll be able to do one entire human genome for $1,000," predicts Rothberg, whose first company, 454 Life Sciences, was the one that sequenced James Watson's genome.
A $500 exome sounds great, but how much data is delivered on each position within all the genes?  Probably enough to genotype normal DNA but if you're analyzing a cancer tumour I'd bet you want to spend more on data that gives a more detail perspective (for various reasons).  If that's the case, a $500 exome might not give you the depth of information really needed to interpret what's gone wrong within the tumour.

Moral of the story: Don't go by cost alone.  Sometimes the cheapest data ends up costing the most in terms of lost opportunities.

Tuesday, June 25, 2013

Tapping Entrepreneurship In Biotech Startups

Gordon Webster, at The Digital Biologist, on Merck Serono:
Rather than re-investing this money in its own, in-house R&D, Merck Serono is instead, seeking to tap the rich vein of entrepreneurial innovation that exists amongst the smaller, more specialized biotech companies and startups in the Boston biotech cluster.

Creating a genuinely entrepreneurial environment within large organizations is challenging, especially given their innately hierarchical management structures. And while the risk associated with such venture-funded biotech investments is significant, it is arguably a more flexible and manageable risk for Merck Serono than expanding its own internal pipeline in such a challenging climate for commercial life science, with all of the concomitatnt overheads and commitments that this would entail.
Though I'm sure entrepreneurial innovation is alive and well in specialized companies, I wouldn't necessarily infer that entrepreneurship can't co-exist with larger organizations.  Sure, there's hierarchy in any organization, but 'intrapreneurship' has been a buzzword for those with a passion for launching something new, but without the circumstances needed to hit it out on their own.

So if you have a great project idea, nurture it. Can you establish a better identity for you or your company?  Rebrand yourself.  Even in academia, intrapreneurship helps create productive and successful faculty.

But I digress; back to startups.

A few months ago, GEN had a great article about 'creative acquisition' as a natural model for biotech (which I expanded upon) describing how innovation gets outsourced to startup companies, particularly in biotech.

And while the main point of compartmentalizing research in startups is to keep focus on a small number of objectives, it's also to place each level of research in the hands of those capable of hitting their goals.

More importantly, it prevents people from working on things they're not good at.

If you're the best at one or two tasks, you're probably average in many others, and if you have to do everything yourself you're probably going to achieve little.  And if you're a Merck shareholder, having your startups learn new things that others already do better isn't just wasteful, it's strategically risky.

Monday, June 24, 2013

Iceland Blocks deCODE Genetics Use of Predicted Genotypes

Jocelyn Kaiser, at Science Insider, on deCODE Genetics:
The current dispute involves an approach in which geneticists calculate the odds of whether an individual carries a particular genetic variant without directly sequencing their DNA. Combining known and estimated genotypes for its research participants with genealogical data, deCODE is estimating what it calls in silico genotypes of close relatives of the research volunteers, essentially giving the company genotypes for all 320,000 Icelanders. The firm then uses these estimated genotypes for individuals as controls in its studies and also combines them with health records for patients who are part of a disease study in Iceland but whose DNA has not been sampled.

Iceland's Data Protection Authority (DPA) ruled that if deCODE wanted to continue this strategy—it asked to impute the genotypes of 280,000 living and dead relatives of the research volunteers and link them to certain hospital records—it must obtain the relatives' informed consent.  
deCODE's CEO, Kari Stefansson, is obviously not content with this decision, especially as the analyses weren't forbidden under previous rules and the company has published several papers using the approach.  In a way, imputing genomic information and using it in place of actual DNA samples of the patients concerned (who have not consented to having their genomes used) is a bit of a loophole. 

A loophole is defined as some areas of rules or laws that isn't clearly defined, and thus permits certain actions that probably wouldn't be permissible with a clearer interpretation of the rules.  Eventually, loopholes get closed, an in nearly all cases leaving someone crying foul.



Thursday, June 20, 2013

Bill Nye the Science Guy Visits Iowa State Campus

John Schwartz at The New York Times wrote a short backgrounder on Bill Nye the Science Guy, based on Nye's recent visit to Iowa State. 

Besides describing Nye's brief work at Boeing (I didn't know about that), Schwartz quotes Nye's brilliant rebuttal against Creationist parents:
“If you want to deny evolution and live in your world — in your world that’s completely inconsistent with everything we observe in the universe — that’s fine. But don’t make your kids do it, because we need them. We need scientifically literate voters and taxpayers for the future.”
And I'd add that we need more scientifically literate legislators and politicians, too.

Wednesday, June 19, 2013

Johnson & Johnson Adds Aragon Pharma Drug to Oncology Drug Portfolio

On Seeking Alpha:
Investors reacted somewhat favorably towards Johnson's acquisition of Aragon Pharmaceuticals. Four years ago the company already acquired the Zytiga cancer drug when it acquired Cougar Biotechnology for $1 billion. The drug has been a huge success, used by more than 60,000 patients. Zytiga generated first quarter revenues of $344 million, up 72% on the year before.
As a drug that inhibits the formation of testostone precursors, Zytiga was well positioned to be of benefit in  prostate cancer.  It looks like Johnson & Johnson is on its way to recoup the outlay for Cougar Biotechnology in the near future.
Still, a potential market entrance of ARN-509 is not likely for a few more years. Unlike Zytiga, ARN-509 has the potential to help as well for patients in whom cancer has not spread yet. Combining both technologies and drugs could ease the pain from competition arriving in 2016 when generic drugs will impact Zytiga's sales.
ARN-509 is an interesting complement to Zytiga because it's not acting through another metabolic pathway; the primary mode of action is to inhibit androgen receptor localization to the nucleus, preventing it from activating downstream genes.
If approved, the potential new ARN-509 drug would have market exclusivity until 2028. Johnson & Johnson is still awaiting late-stage trials to asses the methods of usage of ARN-509, including the possibilities of whether it would be used in combination, or sequentially, with Zytiga.
I'm looking forward to seeing what strategy ends up as the better one.  Combination therapy of Zytiga and ARN-509 might corner cancer cells and kill them effectively, but might also pressure them to evolve to another, perhaps undruggable, state that's not dependent on Androgen Receptor or the formation of testosterone.  On the other hand, sequential therapy with one drug may push the cells to become more sensitive to the other.

This growing prostate cancer portfolio is definitely one to watch.

Tuesday, June 18, 2013

Naturopathic Doctors Convocate At University of Toronto

At Sandwalk, Larry Moran rants a bit about the recent Naturopathic Doctor convocation ceremony at the University of Toronto, in where else, but Convocation Hall:
Convocation Hall is where the university's own graduation ceremonies take place (the name is a clue!). The average person will assume that the University of Toronto supports the Canadian College of Naturopathic Medicine and that our faculty has no problem with graduating quack "doctors." This practice should stop.
The post spawned something of a flame war in the comments, but an exchange between Barry and Larry (That's not a joke) is thought provoking.

Barry commented:
Larry, I'm a little puzzled by your position. I presume (based on your reasoning over the Ball State case) that if you had faculty at the University of Toronto who promoted Naturopathic medicine you would defend this as an "academic freedom" issue. Yet having an independent College of Naturopathic Medicine present degrees there somehow brings the university name into disrepute and you feel sufficiently strongly about this that you say "This practice should stop".

Surely the "offense" in both cases is the presentation of material antithetical to the concept of a university education, which is apparently OK if it is done by your own faculty, but to be prevented if done by an outside body who, I assume, you think more strongly positioned to damage the reputation of the university through association than an assistant professor in their direct employment.
To which Larry replies:
I would strongly oppose the teaching of Naturopathic medicine at the University of Toronto just as I would strongly oppose teaching creationism. I might even picket the classroom or hand out pamphlets to the students.

What I would NOT do is try to get a professor fired or threaten him/her with a lawsuit.

I don't see any inconsistency in advocating that the university not rent out space to a private school that teaches nonsence [sic].
I don't know how much the average person would read into a University renting space to an organization, but I wouldn't take it as a message of support, tacit or not.  Rent is rent, and at the end of the day if the University isn't signing on to the diplomas, it doesn't matter. 

I've never seen a diploma bearing the logo or brand of the location where the degree was granted when it's not actually affiliated with the degree.

Monday, June 17, 2013

The Risk Associated With Living Downwind of Asbestos Sources

Usman Paracha writes about a recent report on mesothelioma risk and cement plants:
Researchers found that the chances of developing environmental pleural mesothelioma was higher in the population living within the “500 m of the [cement] plant than in those living in a radius of 500-2000 m and much higher than those living at 2000-10 000 m.” Moreover, the chances were even higher in the downwind direction from the asbestos source.
The original paper isn't Open Access, but it consists of a single table of observations that show a 161-fold higher rate of mesothelioma, a type of lung cancer, associated with living downwind from a cement plant under study. 

Though it's a one table paper, what the researchers did very nicely was isolate the cement plant as a contaminating source of asbestos by considering people living in proximity of two other asbestos-using industries in the region; a brake factory and an asbestos insulation factory:
In the present study, there were no cases of pleural mesothelioma around two other factories in the area in which asbestos was used. This is not surprising considering that in the asbestos cement plant a large volume of raw, fluffed asbestos was manipulated in the open air, while the other two factories worked with much smaller amounts of processed asbestos in closed areas.
So as they state, it's not a shocking observation, but nevertheless helps make the study a good example of experimental design.



Thursday, June 13, 2013

On Knowing Your Scientific Market

It's sometimes said that there's never a shortage of good ideas, and that's never been truer in science.  Good people, in contrast, are always in short supply, and resources are always limited.

So picking the best ideas to pursue given your constraints of time, people, and resources, is a fine skill to develop, but there's no consensus as to what this skill actually means.  What do you choose to do?

Do you stay true to academic ideals and work on the best scientific questions, regardless of whatever lack of money or talent that restricts you?  Will you hit your goals before someone else (who might be in a better position to do so) does?

Or do you chase the problems that potentially have high impact, even if that means that you end up chasing the hot topics of the day and never become an expert in any one field?  Buying a new hat every day will cost you dearly.

A middle road might be to stay true to your ideals, focus on what you're good at (and can deliver on in a timely manner), and research the market for problems that you, and only you, can solve efficiently.  Think about what you have, what your competition doesn't, then deliver.  No one will stand in your way.

There will always be the big important problems to solve, but they probably won't find you that often.  You have to look for them.

Tuesday, June 11, 2013

Article on Genetic Technology Patenting on Signals Blog

Another one of my short articles has been published on Signals Blog.  

This time, I look into arguments against Vernon Bowman, the farmer in Indiana who cleverly executed a selection experiment for Roundup Ready soybeans and in the process managed to recover Monsanto's products without paying licensing fees.

Unfortunately, it was so clever that he ran afoul of the United States Supreme court and ended up paying over $80 thousand in fines to the company.

As someone biased to the biotechnology industry (I'm basically in it), I have to take the stance that protecting genetic technology patents are necessary for a functioning industry and to encourage new investment and technology development.  Fortunately, that's the same argument the court took against Bowman so I'm not alone in my position.

I encourage you to read the article and share your thoughts either on this site or on Signals.

Monday, June 10, 2013

Canadian Biotech Needs Canadian Philanthropists

Wayne Schnarr posted a great summary of Canadian biotech over at The Cross-Border Biotech Blog:
To paraphrase the conference luncheon panelists, there is no D in R&D unless there is good R, and the best way for government to help fill the pipeline of new technologies and companies is to continue funding basic academic research. If they do not, their short-sighted (to the next election) approach will have a negative long-term impact.
A volatile and difficult economic environment has forced governments to focus on stimulating economic activity and job creation. The complaint about an inability to fill job vacancies has been heard for decades – although the province and the industry may differ. From my perspective, this is a result of either a surprising industry success or poor planning a decade ago by governments and industries. If governments and industries start talking now about the professionals they will need in ten years, this might be enough time for a response from the slow-moving academic institutions, at all levels from high school to graduate school.
At this point, it should be pretty obvious that one of Schnarr's arguments is for long-term, stable, funding needed to build up science at per-commercialization stages. He touches on government as the source.

But is government the right source of this funding?  There's also the message that the election cycle has a disruptive effect on basic science, which suggests that government funding is not the best model for sustainable science funding.

One possible source of less volatile basic research funding are private foundations, but what Canada lacks is one or two organizations on the scale of a Wellcome Trust, a Harvard Endowment, or a Howard Hughes Foundation.

Some might suggest that it's not possible for a country as young, or as small as Canada, but you only have to go as far as finding a few multibillion dollar fortunes to see that Canada does not lack the kind of capacity for wealth creation and aggregation that's required to support big, world scale, revolutionary research that's truly independent from the short term pressure to yield returns on taxpayer investments. 

Now I'm not suggesting that the likes of Galen Weston relinquish his money for science (he does, after all, have a company to run without really having to worry about minority shareholders), but that Canada in aggregate certainly has a few billion to spare.

So I tend to lean towards the idea of good entrepreneurial foundations as the big funders of science over hospital based foundations, which usually have terrible expense ratios, or government grants, which are volatile.

Here's the final challenge: Find the few people in Canada that not only have the scale of funds to set this kind of research in motion but who are also farsighted enough to notice that they're the only ones with the ability to do so.  Launching up smart institutes using private funds is only just beginning.

Tuesday, June 4, 2013

How To Make a Math Talk Comprehensible, From A Biologist's View

Kresimir Josic muses about how mathematics research talks are much less comprehensible than biology research talks:
The reason for the differences may be that good presentations are much more important in biology. Even a mediocre talk will raise eyebrows, and it can kill your chances of getting a job. And a poorly written grant will not be funded, no matter how good the ideas. In biology there is a high overlap between people that do excellent research and give excellent presentations. Less so in mathematics. The cynical reply here is that these are just better salesman and get more funding, and hence run bigger and more productive labs. But perhaps these are simply the people who view the presentation of their research as an integral part of their work.
I completely agree with Josic that presentations and communications are central to biological research, but here's why I think that is.

I'm not too familiar with the breadth of research subjects in mathematics, but in biological research, the topics are so diverse that you're pretty much guaranteed to be presenting to audiences where 90% of the members only have general grasp of your specific field, let alone the precise topic. 

Thus, every presentation is a general primer on your topic and a lead-in to the problem you're trying to solve, why it is important to the field, and why it should be important to them.

The second aspect of learning how to present biological research is picked up in grad school; tailoring your presentations to your audience. 

Given the same set of data, a presentation will be set up differently depending on whether it's a lab meeting, an advisory committee meeting, or a departmental meeting.  Each of these groups has a specific need from you, which you must determine in advance and plan your presentation accordingly.

There are a ton of books on how to give great presentations:  I liked Presentation Zen for it's reminders to keep slide clutter to a minimum; The Cognitive Style of Powerpoint by Edward Tufte for understanding visual communication; and more recently How To Give a TED Talk for advice on putting impact into your message.

Lastly, I wouldn't say calling biologists better salesmen is cynical; it's partly true.  There's no shame in selling your science if you're trying to share a message with your audience (overselling your science is a different story).  There's a little bit of salesmanship in every aspect of life, and research is no different.

Monday, June 3, 2013

Classic Science Problem: Information Hoarding

Here's a bit of wisdom on information hoarding from Seth Godin:
If your project or organization depends on knowing things that other people don't know (but could find out if they wanted to), your days are probably numbered. 
Information hoarding isn't a problem unique to travel agents or real estate agents, it's actually perfectly applicable to science too.  If your groups' competitive advantage comes solely from restricting access to data, analysis results, clinical samples, it might last for a while but eventually that advantage will erode. 

In the research gift economy, it also means that if others know of your data hoarding practices they'll likely reciprocate in turn.  What goes around comes around.

Lastly, the advantage will disappear especially quickly if the information is easily replicable, like contracting out data generation on commodity samples like human cells or tissues. Godin's penultimate paragraph notes this concisely:
Information is in a hurry to flow, and if someone comes up with a better, more direct, faster and cheaper way for information to get from one place to another, they will eliminate your reason for being.
With the advantage gone, and that remains is negative sentiment.