Search This Blog

Wednesday, April 17, 2013

Euclids Elements

In 'Euclids Window' by Leonard Mlodinow we are taken through the history of geometry. It's not at all boring as some might assume. There is a history lesson on Alexandria, the leaders, and the library for example. It takes the imagination to a time where information was kept precariously on rolled up sheets of papyrus. Alexander the Great borrowed such texts from other leaders and didn't give them back! Imagine how much easier our own lives are now yet how much lazier that has made us. In the days before the 1000 year dark age, geek knowledge was in fashion. The math employed by the Egyptians was a commodity that the Greeks would travel by wooden ships to obtain. This made the information physically valuable to collectors of all things foreign. Think of those individuals as venture capitalists. They know something is valuable by they know not why. It was intellectually valuable to Euclid and the scientists he influenced. They knew why.

Euclids, 'Elements', his seminal work on Geometry, relied on the accumulated information on the subject. It is believed that most of the theorems were proven by other mathematicians. Perhaps the first sign of a powerful new scientific understanding would be a tall building or a way of having water available for farming. Egypt for example, had the pyramids. Someone capable of understanding the science and engineering had to grab some papyrus and go find out how they built them. It took Euclid many years to collect information from these kinds of writings and finally sit down to begin his work. He laid out his lifetime of interest in geometry, the things he felt were powerful in 13 rolls of papyrus.

This collection, 'Euclids Elements' has influenced people in ways beyond geometry. Bertrand Russell, for example, adopted the axiomatized deductive structures that Euclid's work introduced. (I stole "axiomatized deductive structures" from Wikipedia.) It is important to understand that sentence though. Axiom: A starting point of reasoning. Deductive reasoning links premises with conclusions. The structure of Elements is genius. 

Let's apply some "axiomatized deductive structure" to the Cargo Cults of biotechnology. There is much that we have learned since 1980 regarding the life sciences. As far as biotechnology and medical science are concerned however, we have created a huge steaming pile of information. It is not easy to find the starting point of much of the research then follow it logically to the conclusions that have been set in stone. I say "set in stone" because that is how biotechnology science begins. Let's look again at On-Q-ity. 

A group of investors will put money up to advance a science/technology towards a profitable healthcare product or service. On-Q-ity diagnosed cancer using biomarkers. After the money is in place PhDs are hired to instruct bachelor degreed people on how to "prove" the science and technology are useful. The conclusions are set in stone. Now imagine the new hires finding out that the science/technology isn't panning out as published? It doesn't work the way Bruce Booth and co-workers believed. Why did the investors believe that it worked? 

It is in this area, where the people with power run into trouble with the real boss of us all. Science doesn't care about a Venture Capitalists hopes and dreams. This area, science, is where the poor bastards with new PhDs and bachelor degrees have to work. It is where I worked and one day had to start this blog to deal with the insanity. Where a man who lived around 300 BC was once able to accumulate knowledge and put it down for people to learn from for over two millennia, we have arrived at a place where science is taken as a secondary function of a science based endeavor. The analysis of Bruce Booth on why On-Q-ity failed was an "Elements" of its own. It offered up the view from Cargo Cult leadership on what matters. The reasons for jumping in were not scientific. The reasons for failing were. 

Will this simple message ever make its way to the world? Science matters.

One axiom from my own axiomatized deductive structure before I end. Biotechnology in the pharmaceutical industry is plagued by unsubstantiated (non-reproduced) published research. At the core of both the bad science and the bad investment is the desire for money and power. A career, a new car, a busy schedule for speaking engagements... The signs of a true success. They are all very nice but only a true scientist would shun them all if it meant being a bullshitter. Perhaps among the ranks of our janitors and farmers are the scientific sages of our modern world. Our brightest minds are lying dormant until the day comes when pure thought returns to fashion. 

Tuesday, April 16, 2013

Negative Results On-Q-ity

Bruce Booth has been an interesting blogger on the topic of biotechnology. He is the recovering scientist turned early stage VC. I've made the analogy that he is like the recovering gambler turned gambling advisor. You take the risks, I'll mitigate mine by using your money. Bruce wrote this post about a company he and his co-workers backed with other people's money. It failed and the post mortem was indeed the product of a most clever wordsmith.

Bruce's original thoughts on the investment opportunity:

  • “Hot space at nexus of two exciting themes in cancer and personalized diagnostics”
  • "Strong team with a track record in building diagnostic businesses"
  • “Robust financials with attractive recapitalization” 
  • “Very strong intellectual property”
  • Perception of “multiple possible exit paths via either diagnostic or oncology-focused pharma M&A or possible IPO”

Sexy, sexy, sexy!!! This company had a hot space, strong team, robust financials, strong IP and an exit door. What would the Reproducibility Initiative have added to this assessment? What Bruce was buying with other peoples money was a service. The money maker was suppose to be a cancer diagnostic service. Where in the above analysis do we hear anything about the science?

We here at the CCS believe that the number one reason most companies fail, and most projects in any company (successful or not) also fail is due to Cargo Cult Science. Whatever the reasons behind practicing CC science, it will not pan out in the end. The cargo planes will not land. In spite of this fact, money can still be made. This is what motivates Bruce to get the hell out of science to become an investor. I don't blame him, I got out too. The only difference is that I wouldn't waste any money on any biotechnology unless the technology is based on sound science. From that point the technology must then work as advertised. From that point you can start to look into the hot spaces where a strong team and IP can... MAKE MONEY!

After all... it is a business. Here is an easier decision, one would hope. The Office of Research Integrity (ORI) ruled that Andrew Aprikyan committed scientific misconduct. He no longer works at UW. Where does he work? He works at Stemgenics Inc. Stemgenics Inc. has recieved $180,000 USD in the form of a small business loan titled, SBIR Phase I: Membrane-Permeable Nano-Materials for Cell Reprogramming Solutions. The likes of Bruce Booth would probably have to pass in Stemgenics Inc. but the NSF likes what they see. In time Andrew Aprikyan could pull the wool over the diligent eyes guarding VC capital, except for the strong management team. Having a founder who has been ruled against in a misconduct case at the ORI should be a red flag for a venture capitalist. You may be able to hide this fact, but is the scientific foundation equally as shaky as the founder?

Ultimately, you can analyze the founders and the company all you want and you can easily fail. A better analysis would be focused on the science. Rather than doing what Atlas Venture did, hire a team at The Reproducibility Initiative to look into the science behind the technology. Hire anyone who has a laboratory and the ability to test the theory. If you think venture capitalists have a scientific method for picking winners, you would be mistaken. Science often takes off on paths that lead nowhere. That is part of the fun. The question for investors is "When?" is a business plan ready, from the scientific perspective? Without reproducibility, you are going to do what Atlas Ventures and On-Q-ity did. You will spend millions and millions more having the new lab staff try and validate the science. Chances are they will fail and your money is gone. 



Monday, April 08, 2013

Positive Results

I'm obsessed with the Amgen Study. It is true that the results are what I expect thus my own bias is satisfied. I don't like the results, but they support my belief in this whole cargo cult idea. So I went to set up the link on a recent post and I spent the time to read some of the comments. The first comment I read mirrors my thoughts.

Publication bias is a problem in all fields of research. The results of a paper should actually receive zero weight in the evaluation of its quality, otherwise there is the motivation to cherry-pick the data that give the most impressive result. The measure of quality should be the way the results were obtained – size of sample, experimental procedure, endpoints used. Ideally the reviewers of a paper should not see its results at all, only the description of the experiment.
But then there are a number of comments highly critical of the paper. What did these people not like?
The claims presented here are pretty outlandish. Particularly relevant to "Hematology and Oncology" we now know that mice housed under different conditions with different microflora can have vastly different outcomes in any model, not just cancer. To suggest academic incompetence or outright unethical behavior is offensive, and is a particularly narrow view of why experiments are difficult to reproduce.
This commenter seems to be defending non-reproducible results. Scrolling down we get a comment on this comment.
Reproducibility is the crucial part of scientific method. If indeed "mice housed under different conditions with different microflora can have vastly different outcomes in any model" (quot. from one of the comments above) then I wonder if mice presents a useful model-study that can be used outside a particular lab and on which further investigations can be built upon outside that one lab. 

Perhaps the commenter critical of the Amgen Study does not have a firm grasp on what makes for good science. The natural urge to simply defend could come from the human condition of favoring positive outcomes. As the first commenter points out, results shouldn't matter. That is a very interesting concept that not everyone will understand.

But all of this leads me to the latest TED Talk from Ben Goldacre. He too brings up the Amgen Study. He talks about this idea of positive results and our tendency, as humans, to favor them even when negative results are the accurate results. Hiding them leads to trouble, scientifically speaking.




Again, Feynman CCS

One example of the principle is this: If you've made up your mind
to test a theory, or you want to explain some idea, you should
always decide to publish it whichever way it comes out. If we only
publish results of a certain kind, we can make the argument look
good. We must publish both kinds of results.
 
I say that's also important in giving certain types of government
advice. Supposing a senator asked you for advice about whether
drilling a hole should be done in his state; and you decide it
would be better in some other state. If you don't publish such a
result, it seems to me you're not giving scientific advice. You're
being used. If your answer happens to come out in the direction the
government or the politicians like, they can use it as an argument
in their favor; if it comes out the other way, they don't publish
it at all. That's not giving scientific advice.


Thursday, April 04, 2013

Amylins Fire Burns Out

Amylin Pharmaceuticals was a superstar biotech in San Diego. By the time Bristol Meyer Squibb acquired them they had over 1200 employees. The decision to set most of them free is a typical end to such an acquisition of a biotechnology company. The few that actually produce a product usually get acquired.  Those at the top get rich and the workers who contributed to the success get their walking papers. 

In the case of Amylin we have an interesting product. According to Consumer News:
If you take Byetta or Januvia to control your blood sugar levels, you might have a small, but increased risk for a condition marked by stomach pain and nausea. Not only are the drugs more expensive and no more effective than our Best Buy Drug picks for diabetes, they're associated with an increased risk for pancreatitis, or an inflamed pancreas, according to a new study in JAMA Internal Medicine. We recommend asking your doctor if you shouldn't be taking another drug in the first place.
There is, of course, a difference of opinion among the executives of BMS and Amylin. BMS paid billions for Amylin because Exendin was a sexy new alternative to the usual type II diabetes drugs. The disease itself can be easily reversed by a change in diet and proper exercise. The pharmaceutical industry has long been making bets that sick Americans would prefer a pill, and its side effects, to changing what they eat and how much exercise the get. Even if the pill doesn't offer improved results from cheaper drugs, they still have a sales force and lazy greedy doctors to offset the shortcomings.

Amylin certainly had its experts in the area of type II diabetes drug development. Imagine a court of law where there is a case known as Type II diabetes drugs vs Diet and Exercise. The case would demonstrate that pharmaceutical experts do not actually study the disease. They study the disease endpoints that they set out to alter for a more tolerable existence for those who choose to remain ill. The diet and exercise experts would offer their evidence that the disease state is caused by bad decisions. After an individual makes poor diet decisions  and get themself into the disease state, the pharmaceutical industry keeps them there. They also provide new problems. The pharmaceutical experts have also been caught hiding data from the FDA about hearts risks from Byetta and stalling the agency's efforts to learn more about the risk.

The hundreds of people who will be laid off as a result of BMS's decision to shut down the San Diego operation are mostly tribesmen. They did their job, looked to the sky and lo and behold something appeared. It needed help coming across as cargo but the leaders convinced the government that it was cargo. Once the cargo was given the official seal of approval, the cargo cult was made legitimate. It was sold, the cargo was secured, and the tribesmen are free to go find new cults.

“It’s not companies that matter, it’s people that matter,” said Walshok, associate vice chancellor for public programs and dean of UCSD Extension. “San Diego’s strength is in the talent that is resident here.”
If only BMS felt the same. The people who matter will spend a lot of time on unemployment. Some will switch careers. The expertise in type II diabetes research? Did it ever exist?




Wednesday, April 03, 2013

Award Winning Excellence

When we want more than we can actually obtain, we find substitute solutions just like the Cargo Cults. We have religion to help us deal with our existence and mortality. We have exalted leaders to give us a sense of order. We have corporations to keep us employed. And yet, the limits of these human constructs continue to let us down. The cargo doesn't always come. Only the laws of science seem to be constant.

At the Oscars, when a celebrity has to step out for a smoke or a bathroom break, their are people who take their seats so as to mask the stars absence. When we have an area of life that lacks a scientific explanation, we have seat fillers. In the world of professional scientists (I didn't say science), the truth is not coming to us as fast as our publications are being churned out. The area between science and what we want to believe is where we can put ourselves on the right or wrong path. It's the area where we went from assuming a god created us to beginning to understand genes and the structure of DNA. It's also the area where Alchemy exists. There is a fork in that road where scientific leadership is suppose to lead us. When the leaders decide that RNA interference, for example, is the right road we end up with years research and billions of research dollars wasted. We wanted our scientists to provide us with a tool to make future research easier. We got it.

Here is an example of what I am talking about, taken from the world of education. We as human beings want our leaders to make us smarter. We have ignored our limitations and we've given the leaders an impossible job. What do they do with their task? They succeed!

Three Dozen Indicted in Atlanta Cheating Scandal!
...the teacher came around offering information and asking the students to rewrite their answers. Juwanna rejected the help. According to Howard, Juwanna said that when she declined her teacher's offer, the teacher responded that she was just trying to help her students. Her class ended up getting some of the highest scores in the school and won a trophy for their work. Juwanna felt guilty but didn't tell anyone about her class' cheating because she was afraid of retaliation and feared her teacher would lose her job.
We wanted so badly to see the students get better grades that we created leaders to get the job done. Failure was not an option. We handed out medals for success and pink slips for failure. Then we had to face reality. We had exceeded the limits of human capabilities. The humans had to cheat or try their best and hope that things turn out in their favor. When your livelihood is riding on the outcome it is not out the question for otherwise honest people to cheat. We must also assume that the most dishonest among us will do the best. Juwannas teacher certainly did well. She won a trophy. Dr. Beverly Hall, the schools superintendent appears to have won a Nobel Prize! Just look at that medal.

In education we have a flawed system to begin with. We begin with the assumption of perfection, 100% or 4.0. Each time a student fails to tell us what we told them to think, we dock points from their perfect grade. The educators indicted in Atlanta took were suppose to bring their students average closer to perfection. They found a way.

We have to be careful what we ask our leaders to do for us. There is no shortage of highly skilled people willing to tell us what we want to hear.