Search This Blog

Thursday, July 26, 2012

A LinkedIn Conversation

Modern technology has provided us with the ability to offer up our opinions. If you want to get into an argument online just offer up your opinion and wait. If you are on LinkedIn however, you are engaged in a professional conversation, you are not anonymous and you don't want future potential employers to see you offer up an honest but surly opinion. You end up with a Cargo Cult-like conversation.

The topic is phage display ELISAs. The worker used up his protein target panning for binding phage. Now he wants to use suspension cells expressing his target as his target in the ELISA. The first question one has to ask is whether or not it would be easier to find more of the pure protein than to develop an ELISA with suspension cells as the target.

There are many suggestions that are at the level of solving the problem stated. Those who wish to be leaders will accept that problem. It allows them to make suggestions but still leave the final decision to the worker. The probability of failure is high, but the accountability is not on the leader. A true scientific leader however, is the one who asks the questions. The scientist asks the questions because the scientist best knows what the problem is.

I've stated often that our education system selects for the best question answerers. The question askers often come across as the students. They seem naive, in need of a teacher to answer them. A scientist is someone who has the intellectual confidence to ask questions that many might consider to already have an adequate answer, and that answer has been placed into the heads of the educated. As is the case in the above problem, the intelligent response on this public forum would be to answer the question. Be the teacher to the student. Yet I think there is another approach which would have a higher probability of getting to the heart of the matter. Ask why the question has been posed. Question the question. The worker is trying to verify binding of phage to target. Has he set up the best experiment to answer his question. 

Ultimately, this is not a phage display problem. It is an ELISA problem. There is no phage problem because we have a phage expert in our lab with all of the skills needed to do the job. He is missing a reagent and we need to get him a substitute. That is our job as leaders accountable for the successful completion of this task.

As leaders we must take into account time and resources. We must also ensure that we get the best answer to the question. The most  important part of developing any ELISA is knowing the difference between signal and noise. Phage ELISAs have less separation between signal and noise than ELISAs with purified proteins and antibodies. With this new twist of using suspension cells in place of purified protein, one has to wonder what the signal and noise will be once you fully saturate the surface of your ELISA plate with your cells. Will you be able to distinguish the difference between noise and signal with this system?

Verification of binding is the forest. Attaching these cells to a plate is the trees. Make a list of your ideas and the suggestions you get from colleagues, use design of experiment to most efficiently run a set or series of tests and knock it all out at once. If nothing works, quickly move on to trying to get more pure protein.


Tuesday, July 24, 2012

The Five Year Outlook


Outlook for the next 5 years in drug innovation

Roy Berggren, Martin Møller, Rachel Moss, Pawel Poda & Katarzyna Smietana
The decade-long crisis in the productivity of pharmaceutical research and development (R&D) has been widely discussed1, 2. Indeed, expectations of reduced returns on R&D investment have led large companies to scale back their R&D substantially in recent years, and there have also been shifts in R&D investment and activity in particular therapeutic areas by the industry overall.

Nature Reviews Drug Discovery 

Decade long crisis?

What is the value of any five year plan or outlook? No battle plan survives contact with the enemy. The intellectual firepower of this article, likewise, will have to do battle with the reality of an industry that operates randomly. Without control over the randomness of FDA approvals combined with the lack of systematic training and educations that create researchers and research leaders, there is no way to predict our future. Perhaps a better approach would be to look at the 90% false findings in published research and attempt to predict if that will change.


The original assumption in predicting where we will be in five years, is the assumption that we know where we are right now. The assumption that we are in the midst of a ten year crisis is conjecture. The assumption that we can predict the future is just silly. The most scientific thing we could do at this stage of the drug industry would be to define where we are, why we succeed and fail, and what we can do to weed out more of the Cargo Cult. 

Monday, July 23, 2012

More Than a Bachelor Degree

I recently read that higher degrees in computer science are not all that desirable. The skills needed are learned during an undergrad education and mastered on the job.

I was talking with a couple Pfizer scientists. They had a task that they were going to hire a temp to tackle. They needed that person to develop a method to grow their virus in a particular cell line. They went on and on about the virus, the cell line and how the virus was going to be used. They had their narrative down to... a science. So much so that they got the okay to go out and find the temp to do this job. There was two things missing in their narrative regarding the temp. They failed to say "upstream process development" or "design of experiment".

During this talk the Pfizer scientists mentioned that they had very little written information. The new kid was suppose to show up, read their minds, and complete their project. Ultimately, his/her job would be to figure things out and train his/her superiors how to talk about his/her work. There would be a technology transfer, from the kid, to the people who need to know.

One of the reasons biotech/pharma and science in general has had such a dismal record lately is our faith. We have faith in the PhD. We have faith that the university is spitting out highly skilled minds. We have faith that a PhD is THE prerequisite for getting the job done. In the Pfizer example however, they have no intention of getting the job done. They will sit there like their advisory committee passing judgement, but not designing proper experiments.

What is the purpose of the PhD if they are not the ones designing the experiments? They certainly do not work in the laboratories. When a new biotech/pharma scientist is hired, is it assumed that they know about upstream versus downstream processes? Who trained them to be in the business? Who trained them to be leaders? Do they know how to put together a team? Can any of this be sussed out in a conversation or two that is had during the hiring process? I say no, and the only way to ensure that the PhD gets the additional training is to actually train them.

I give these Pfizer scientists little chance at creating an upstream process for their virus project. They plan on hiring a junior scientist to do all of the work. They have done a fine job of describing the Cargo that they hope for. They have pointed to the sky and explained that the Cargo will come from there. The only thing left is to find someone who knows the proper ceremony to perform. They'll know if he/she succeeds by staring at the sky and waiting.