Transcript dougherty informs kl..

Exploring Unique Learning in BioPharmaceutical Innovation
Deborah Dougherty
Danielle Dunne
Rutgers U
Grounded
theory building:
core processes, activities in this kind of innovation, and
how to organize, manage
Focus
on drug discovery, early development (first 6 yrs of
12-16 yr process)
Trouble
Existing
in bio-pharma city!
models little help (feel huge voids, not just gaps)!
Limits of what we know

Technology: presumes linear, decomposable,
scalable, path dependent, Newtonian (laws of
physics), concrete, objective
– Stage-gate, platforms, architectures, modularity, CE
– Engineering based

Life sciences (pharma): non-linear, non-

Others that do not fit: “bio” plus, “nano”
decomposable, non-scalable, pathless, obey laws
of life and social sciences; knowledge is
descriptive, limited, requires multiple levels
plus, health care, ecologies and climate changes,
terrorist and intelligence systems
Focus today
 Techno-hype:
bio-pharma involves scientific core,
multiple sciences and multiple
technologies
 Non-decomposability
 Non-scalability
Techno-hype
Our business is very different from making cars. You make a car, you look at
all the pieces and you know how to optimize the tires and how to put
together the gears… You know in much more detail what you need to do.
Here, it is very (uncertain); given a negative or a positive experiment it
takes still a lot of intuition to make the next one… we value technologies
that vastly accelerate things but we are still dealing with a very complex
system. So imagine that you are in a race where you have this fancy suit
where you can run 10X faster, but if you don’t know where you are going
it does not make a difference how fast you got there
The classical project management comes form engineers and people who are
more attuned to set processes… and processes that are not so much in
the state of flux as drug development. Every project is so unique even
within the same therapeutic area… there are so many caveats and so
many nuances to any… drug program
A biological system is the ultimate in engineering. It has had billions of years
of tweaking to get here…. The tools are different, the methods are
different (from engineering), but still one thing has to work in concert with
another thing to get a function. We just have many more regulatory
levels, most things do multiple things at once… in five hundred thousand
years, cars will probably be very biological in the sense of massive
redundancies and systems overlap…
Past 30 years, all about sticking in
technologies

Microbiology: look directly at cell mechanisms
Biotechnology: cloning, create huge quantities of proteins;
gene replacement (nada); protein replacement (a few:
insulin, factor VIII)
Combinatory chemistry
Rational drug design…
Genomics: know all the genes…
High throughput screening: test million at a time…
Robotics
Structural bio; bio-markers; analytical electroscopy
Assays for everything…

OUTPUT dropped in half nonetheless – NOT just technology!








Non-decomposable
Hallmark of industry!
 Massive redundancies, cannot separate
parts or assume known links
 Molecule, protein, cell pathways,
mechanisms disease process, organ
system, body…
 So, what to optimize? How to manage
whole blob over 16 years???


Conceptually differentiate distinct
knowledge systems
– Core: sciences (biochem, phys, micro etc.,
chemistry, physiology, pharmacology…)
– Recognize difference in goals from academic
sciences
– Technologies (instrumentation to collect and
process data for science, industrialization,
search, techniques (delivery etc.),
representations of life system, lines for science
– Strategy: new role? Risk mgt, long term view
– Process mgt: how to enable each plus
combos, oversee processes, progress,
choices…
– Therapy areas? Think is boundary “space”
Then figure out each one’s unique contributions,
mechanisms of knowing, how enable others

Academic Sciences: understandings, new
insights (fragmented, limited integration, lab and person
based?)

Industrial Sciences: what we don’t know:
questions to ask, how, where to search, what to
make of answers? (searching for clues, iterating into
partial wholes, based on entire DDD process)

Technologies: what we do know: search
engines, high throughput science; generates
answers, bundles old stuff new ways, see wholes, “verify”
certain options
 Strategy??? See what do not know better? Select, mge
categories of risk? Absorb LT insights? Shape exploration
(not exploitation)

Science is about knowing when to stay the course and when to
leave. (how do you know). You know the data to access and the
experiments that have to be done and whether or not they will
cover enough (of the info or situation?) to answer the questions
and test the hypotheses. And your experience with how long it
takes to break through a technical barrier, and whether or not it is
worth your while to stay with it or go over to something else. I
have been in places when I was outlining the constraints a project
faced, getting ready to drop it formally, and then we have a
breakthrough… (referred to picture on the wall of a cover of
Science with her and team…) We wanted to try and find other
molecules like that but they must be smaller… The natural ligand
is much larger… We had some clues that we thought would be
important. There were other receptors in the same family, and
other clues about what could make a connection. Molecules are
bumpier than a flat surface… We visited another pharmaceutical
company and different biotech companies, and even within our
own company. We proved we should do it. We have a
collaboration with a biotech start-up… looking through libraries
with millions of possible ways of how to arrange peptides… I
remember very vividly thinking it was a waste of time. Also we
were trolling in a couple of other areas like natural products, antibodies. Once we got the peptide and they combined, we did our
first crystal structure. We worked with XX at Scripts laboratory,
we did that with an academic collaboration.
Some Noodles






Somehow, if we know the connections among the
systems, can “fire” them off, like lightening
strikes, and see the drug possibility in the whole
life system
Will always take enormous judgements, but
better than separate points
Knowledge accumulates within the knowledge
systems, and also somehow in the connectings
This is purely social
Everyone must do own integration
All knowledge systems must translate to others
proactively
Non-Scalable
Due to complexities, each drug is unique
(and they focus in on each, not on
system)
 No standards, core operating principles or
frames to manage
 Knowledge accumulation not automatic,
but inherently limited Projects morph
anyway (like starting to build a car, end up
with a roller skate)
 Cannot just manage “outside”, must
manage internal or emergent dynamics
instead

A few noodles
Manage the questions: good ones to ask
when, why
 Manage total answers, not single inputs
 Develop platforms for connecting
 Cannot surface problems? So surface
problem setting and ongoing iteration
(problems have multiple causes)
 Fractals? But compounds vary?
 Cycles: over 6 years, circles go counter
clockwise from here back to target,
clockwise forward to drug

Back to beginning…


Enabling searching for clues, iteration, sense of the scientist
in concert with other modalities of knowing: what are key
properties?
(Earnst Mayr, 2000, age 90; AIBS distinguished service award,
bioscience v. 50, 10:896)
– The basic philosophy of biology has become quite
different from the classical philosophy of science
(rejecting vitalistic theories and physicalist concepts…)
and their replacement by an acceptance of the
importance of historical narratives, multiple causations,
population thinking, and the greater importance of
concepts than of laws in theory formations…

Biology, he said, is far advanced with basic phenomena, lag
in understanding complex systems. KLIC too!