Transcript Chemical Informatics - Indiana University Bloomington

Thriving on
Information Anxiety
A Survival Guide to the
Knowledge-Value
Revolution
Sam A. Falk Milosevich
Associate Professor
Chemical Informatics
[email protected]
Home Page
• Introduction
• Common Challenges
• Information Anxiety
• Knowledge Value Revolution
WARNING: this presentation is
rated “G” for “General” Session
G
Technology is the conceptual bridge
linking the operating rules of science and the universe
with those of economics.
— Stan Davis, Future Perfect
• Can you imagine?
– “glass house” turned inside-out
» systems - Twilight of Sovereignty
» simulations - Information Anxiety
» solutions - Knowledge-Value Revolution
– integrated value chain
» platforms with the power
» applications
» people with the problems
– unique value added to the enterprise
» multi-dimensional, parallel
» the next generation, not the next iteration
Future Shock is
the dizzying disorientation brought on
by the premature arrival of the future
— A. Toffler
• How would you like?
– add unique value to your enterprise
» end-user solutions
» enhanced products
» improved business
(authorship)
(publication)
(impact)
– imagination, innovation, renaissance reality
» data acquisition & delivery
(commodity)
» information development
(unique query)
» knowledge-value decision-making ("inform")
– example
» context specific to drug discovery at major pharma
» concept general to all sizes of organizations
Nonlinearity means that
the act of playing the game
has a way of changing the rules.
— James Gleick, Chaos
• Common Challenges
computer-based applications have been
narrowly focused, often under-valued, rarely
become part of the routine; vs. bioinformatics
• Information Anxiety
mismatch between volume of data generated
and dearth of understanding derived from it
• Knowledge-Value Revolution
in post-industrial society, value is more
subjective -- and unique value added is a
matter unique to each consumer
Eli Lilly and Company is
a global research-based pharmaceutical corporation
... working to create and deliver superior health care solutions
that provide customers worldwide
with optimal clinical and economic outcomes.
— http://www.lilly.com
Research is
the heart of the business,
the soul of the enterprise.
— Eli Lilly , 1947
cheaper
faster
better
value:
competitive advantage
classical values
speed:
geographic areas where
can increase market/share
quality:
therapeutic areas where
are or can become a leader
• Cost of developing new drug - $359M
• Need peak worldwide sales of $1B
• Delay in diminishing peak sales:
–
–
–
–
Compound
Synthesis
~10,000
Compounds
Project
Team
First
Human
1,000
8
$31.69/second
$1,900/minute
$114,000/hour
$2,740,000/day
Global
Registrations
3
Clinical Trials
Product
Approvals
Market
Introduction
1 Product
The discovery
is made
with tears and sweat
(at any rate, with a
good deal of
bad language)
by people who are
constantly getting
the wrong answer.
-- J. Bronowski
Drug Discovery
What to look for
• TARGET IDENTIFICATION
How to look
• ASSAYS / SCREENING
- Biology / physiology
- Precise
- Pathophysiology
- Specific
- High Throughput
Where to look
•MOLECULAR DIVERSITY
- Natural Products
- Proteins
- Corporate Libraries
- Combinatorial Libraries
- Automated
• OPTIMIZATION OF LEADS
- Structure-Assisted
Drug Discovery
- Structure Activity
Optimization
from M.F.Haslanger
Natural science does not simply describe and explain nature;
it is a part of the interplay between nature and ourselves;
it describes nature as exposed to our method of questioning.
— Werner Heisenberg
What If?
"If...Then!"
explore
visual model
XY123
What?
Why?
describe
database
explain
rules
Much of the success of modern science and engineering
is based upon our ability to create an abstract mapping
between motions of matter and symbols on paper.
— Larry Smarr, NCSA
Energy = Bonds + Angles + Dihedral Angles + "Improper"
Dihedrals + Non-bonds
X
Total Energy =
C
R

bond


angle


KR (R-R 0 ) 2 +
Y
C
-0
R-R0
Q
K( -0 ) 2 +
C
K [1+cos(n - ) n =1,2,3,4,6]+

Energy Surface
P
X
C
Y
Z
P
QPP
Q
C
R C
Q
C
R
R
X
X
C XY
C
C
Y
ZY
Z
Z
Dynamic Molecules
A

12 r
nonbond ij


Y
Q C
R
K( - 0 ) 2 +
improp
1-4...
C
1+cos(n )
dihedr

Y
B
rij 6

+

qi qj
4 0 rij 2
Human intelligence thrives on context while
computers work on abstract numbers alone.
— A. Penzias, Ideas and Information
It is a persistent mistake to define ‘science’ in terms
of certain features of existing scientific theories.
— John Searle
• Computer Applications
– predominantly computational chemistry
– some computational engineering
• Computational Chemistry
– general molecular modeling
– computer graphics & data visualization
• Drug Design and Discovery
– transform molecular structure w/r
molecular properties; empirical vs. virtual
– mathematical expressions of the laws of
physics are used to model chemical
entities and their transformations; more?
Science is built up of facts, as a house is built of stones;
but an accumulation of facts is no more a science
than a heap of stones is a house.
— Henri Poincarè
• Structure-based drug design goals
– help generate novel ideas for new products
– help compress time for discovery and development
– being first is good enough: typically, Dt = 6 months
• Structure-based drug design results
– some published results
– some clinical trial candidates
– much current work is proprietary
• Structure-based drug design basics
– Combinatorial Chemistry, High Throughput Screens
– Genomics, Proteomics; Clinical Data; Patents
Increasing
Molecular Diversity
• Combinatorial Chemistry: application of
process methodology to repetitive connection
of different building blocks to yield a large array
of diverse molecules.
• Limitations in molecular complexity.
• Rapidly expands compound libraries.
• Challenge is to maximize diversity.
from M.F.Haslanger
Screen Paradigm
Biological Target Selected by Strategy
Screen Development
Screen Validation
Screen Automation and Optimization
Screen Operation
High Throughput:
sensitivity
capacity
Strategy Group Follow up
from M.F.Haslanger
Genomics
• Every new gene discovered
represents a potential diagnostic
or therapeutic target or a drug.
• Integrated genomics tools provide
a means to rapidly validate
potential targets.
Genomics
Genetics
Genomic
Biochemistry
Genome-Based
Screening
Disease Protective Disease Gene Protein-based Small Molecule
Genes
Genes
Pathways
Drug
Drug
from M.F.Haslanger
Gene to Drug
Select Disease
Human
Genetics
Mouse
Genetics
Validated targets
Structural
Char.
Genomic
Biochemistry
Relevant Models
ID Function
HTS screen
Expression
screening
Clinical
Trials
Expression
monitoring
Drug Candidate
Patient
subsetting
Efficacy Decision
from M.F.Haslanger
Knowledge is Powerful Medicine
— Eli Lilly and Company, 1995
Therapeutic
Target
Bioinformatics
Genomics
Pharmaceutical
Lead
Product Developoment,
Submission, Marketing
Chemical Informatics
Molecular Modeling and
Molecular Diversity
Combinatorial Chemistry
and
High-Throughput Screening
Medical Informatics
Clinical Trials
Pharmacokinetics
Health Informatics
Disease Management
Consumers
Information Anxiety
Information Anxiety
is the black hole
between data and knowledge,
[which] happens when
information doesn’t tell us
what we want or need to know
—R.S.Wurman, Information Anxiety
Science <==> Economics
Technology
is the conceptual bridge
linking the operating rules
of science and the universe
with those of economics.
— S. M. Davis, Future Perfect
Knowledge-Value
What is important for the
production of knowledge-value is
... the knowledge, experience, and
sensitivity to be found among
those engaged in its creation.
— T.Sakaiya, Knowledge-Value Revolution
Economic Reality
We are competing globally
on a cognitive basis....
Our economy today is based
upon what you know.
— Dr. W. Leigh Tompson, 13 Apr 94
Strategy
• Scientific & Economic Innovation
– Quality:Effective products & services
– Speed: Efficient processes & systems
– Value: Competitive advantage
• Unique Value Added
– Standardized Concepts in Customized Contexts
– S.M. Davis, Future Perfect
– Social Subjectivity; Small Venture Business
– T. Sakaiya, The Knowledge-Value Revolution
– Commitment: Persistence of Strategies
– P. Ghemawat, Commitment
ADD UNIQUE VALUE
change is resisted when
apparently irrelevant or
out of control
UNIQUE
VALUE
ADDED
.
RE
AL
Y D
IT
IL E E
AB P
LI Y-S
IT
QU
CE
N
A
V
ER
E
M
L
RE USTO
C
TO
Add Unique Value to Compete in
the Knowledge-Value Revolution
UNIQUE VALUE ADDED
MASS CUSTOMIZATION:
• standardize the concept
• customize the context
0
efficiency is a ratio;
effectiveness is not.
DO IT BETTER,
DO IT DIFFERENTLY,
OR STOP DOING IT.
CUSTOMER
RELEVANCE
Put the SCIENTIST
into the SCIENCE.
Enhance personal Creativity
and group Communications.
XY123
"What If?"
"If...Then!"
Supercomputing admits the very large,
the very detailed, the very urgent.
–Boyd & Milosevich, Persp. in Drug Disc. & Des. 1 (1993) 345
m
an
y
large
High Performance (Proc’s)
long
High Performance (People)
huh?
aha!
oh?
yes!
perception
logic
patterns
language
in-sight
active info system
knowledge
hind-sight (20/20)
passive info system
data
“DO GOOD SCIENCE”
SCIENCE
• problem-solving method integrators
• knowledge management systems
APPLICATIONS
• science & technology designers
• information exchange systems
COMPUTERS
• numerics & graphics performance
• data distributed computation
problem-solving focus
Complexity + Contiguity
future
present
Cray-2
past
Scientists need to stay in touch with their
science experiments and with current science.
Simple Rules -->
Complex Results
Chemical Informatics
•provide timely solutions to scientists' problems
•enhance chemists' ability to use all available data
•enable time compression in R&D efforts
•differentiate & support unique capabilities
•integrate and cooperate w/corporate infrastructure
•organize for science function, not system vendor
•educate and consult with scientists on capabilities
•educate and consult with management on costs
•evaluate evolving computer science & technology
It is amazing what you learn if you
take the time to talk to someone.
— M.Jackman, Star Teams - Key Players
Solutions to fit Problems
sys
experts
applications
chemists
hi ear chical pyr ami d
syst ems i ntegr at or s:
k now l edge dom ai n exper ts:
appl i cat i ons devel oper s:
computat i onal & l ab chemi st s:
computer per spect i ve
m et hod per specti v e
appl i cat i on per specti v e
pr obl em per spect i v e
" key" stone ar ches
each mem ber : pr obl em per spect i v e
each ar ch: di ff er ent pr obl em
Next Steps
• Understanding the unique and fundamentally
complex nature of the data, processes, and
problems that characterize the domain
• New acquisition and integration, analysis and
synthesis, or dissemination and use of data
• Technological and infrastructure approaches
to supporting meaningful, long-term
interdisciplinary collaborations specifically
for chemical informatics research
Next Steps
• Integrating strategic technologies for the
internet with a focus on quick impact
– usable and widely deployable networking applications that
promote collaborative research and information sharing.
• Integrating strategic techniques for
pervasive computing and distributed
terascale facilities
– new algorithms, data structures, advanced system
software, distributed access to very large data archives,
sophisticated information mining and visualization
techniques, and collaborative environments for data
exploration and analysis
Next Steps
• Providing innovative educational activities
at the undergraduate level by the transfer of
research results into the undergraduate
curriculum.
• Enhancing inter-disciplinary insights
through collaboration among IT and science
professionals in industry and academia.
• Increasing policy-makers' awareness of the
return on investment in chemical
informatics
A Functional Organization:
• is distinguished by computational science
capability, not computer system vendors
• gives scientists more responsive power (and
more responsibility) on their desktops
• focuses on solving scientists' problems in a
constantly-changing environment
• replaces Mainframe class homogeneity with
Personal Computer style individual diversity
Making the Most of It
– development depends not so much on finding
optimal combinations for given resources and
factors of production as on calling forth and
enlisting for development purposes resources
which are hidden, scattered or badly utilized
– The Strategy of Economic Development in
A.O. Hirschman, Exit, Voice, and Loyalty
– Vertical thinking is digging the same hole deeper;
lateral thinking is trying again elsewhere.
– E. deBono, New Think
–Nature has no outline; imagination has.
– Blake
Why Change?
The system will always be defended
by those countless people who
have enough intellect to defend
but not quite enough to innovate.
...
Politically, change forced by a crisis
is much more accpetable because it is
obvious that something must be done - and
surviving a crisis is achievement enough.
— E. deBono, I Am Right - You Are Wrong
Marathon Challenge
We can do it!
Home Page
• Introduction
• Common Challenges
• Information Anxiety
• Knowledge Value Revolution
WARNING: this presentation is
rated “G” for “General” Session
G
Thriving on
Information Anxiety
A Survival Guide to the
Knowledge-Value
Revolution
Sam A. Falk Milosevich
Associate Professor
Chemical Informatics
[email protected]