The NCI Effort to Understand and Manage the Complexities of Cancer: Intersecting Math and Biology

benhurspicyΤεχνίτη Νοημοσύνη και Ρομποτική

7 Νοε 2013 (πριν από 3 χρόνια και 11 μήνες)

100 εμφανίσεις



The NCI Effort to Understand and
Manage the Complexities of Cancer:
Intersecting Math and Biology





MBI Workshop 4 on Cancer Development,
Angiogenesis, Progression, and Invasion.

OSU January 26, 2009


Dan Gallahan, Ph.D.

Deputy Director, DCB

National Cancer Institute



Outline


Our Cancer Challenge


A Systems Disease


New Perspectives


Modeling Cancer


NCI’s efforts
-

Teaming disciplines, ICBP, PS
-
OC


NIH/NCI funding challenges and opportunities



The Human and Economic Burden of Cancer

21.9

180.7

48.1

586.8

193.9

53.3

190.1

231.5

0

100

200

300

400

500

600

Heart

Diseases

Cerebrovascular

Diseases

Pneumonia/

Influenza

Cancer

1950

2003

Death Rate Per
100,000

More Progress is Needed to Reduce Death Rates

Source for 2005 deaths and diagnoses: American Cancer Society (ACS) 2005 Cancer Facts

& Figures; Atlanta, Georgia

Source for 2003 age
-
adjusted death rate: National Center for Health Statistics, U.S. Department of Health and Human Services,

NCHS Public
-
use file for 2003 deaths.


570,580 Americans will die of cancer this year


1,372,910 Americans will hear the words “you have cancer…” this year


$189 billion per year on healthcare costs


for cancer alone

98
-

Asian Women

331
-

African Amer. Men

A National and International Imperative to Eliminate
the Cancer Burden

Canada

138,000
/

66,000

United States
of America

1.4M
/

566,000

Australia

86,000
/

37,000

China

2.2M
/

1.6M

Austria

37,000
/

19,000

France

269,000
/

149,000

Germany

408,000
/

218,000

Switzerland

35,000
/

17,000

Iceland

1,000
/

500

Ireland

13,000
/

8,000

Japan

521,000
/

311,000

Korea

109,000
/

62,000

Norway

21,000
/

11,000

Estonia

5,000
/

3,000

Republic of
Singapore

10,000
/

6,000

Sweden

43,000
/

22,000

United Kingdom

277,000
/

156,000

Source: Derived from International Agency for Research on Cancer, GLOBOCAN 2002 database

Cancer

Incidence
/
Mortality
per year

7.6 million

people

died of cancer

in 2005


“Cancer is a leading cause of death worldwide and
the total number of cases globally is increasing.”

1990

1995

2000

2005

225

220

215

210

205

200

195

190

185

180

175

Rate per 100,000 (NCI)

Cancer Death Rates

Number of cancer deaths (ACS)

11,941,043 survivors in 2007

Estimated Number of Cancer Survivors in the
United States

Cancer Burden Will Increase

as Baby Boomers Age

Estimated Projections of New Cancer Cases, United
States

Sources:

Cancer Facts and Figures, American Cancer Society, 1997
-
2006.

Jemal A, Siegel R, Ward E, Murray T, Xu J, Smigal C, and Thun MJ.

Cancer Statistics, 2006. CA Cancer J Clin 2006; 56:106
-
130.

Edwards BK, Howe HL, Ries LAG, Thun MJ, Rosenberg HM, Yancik R, Wingo PA, Jernal A, Feigal EG, Annual Report to the Nation on

th
e
Status of Cancer, 1973
-
1999, Featuring Implications of Age and Aging on U.S. Cancer Burden, Cancer; May 15, 2002,

vol 94. no. 10
; pp 2766
-
2792.

2010 2015 2020 2025

2,250,000


2,000,000


1,750,000


1,500,000


1,250,000


1,000,000


0


1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

(in thousands)

Our Cancer Dilemma


More “cured” patients, more high risk patients


Population aging for a disease generally of the aged


Changing Demographics


New biological insights, new concerns


State changes
-

Normal vs. Transformed


Old research models


Current therapies limited


Aggarwal et.al., 2007

Cancer: A Systems Biology Disease


Genes and Genetics


Complex Signaling Networks


Multiple Cellular Processes


Micro
-
Environment


Host Systems


Environmental Factors


Population Dynamics

Time
-

Progression

Histology Variation

Initiation


Progression


Metastasis Recurrence

The Complexity of the
“Interactome”


Protein
-
protein
interactions


Protein
-
DNA interactions


MicroRNA
-
mRNA
interactions

Molecular interaction networks


Growth factors
produced in
adjacent cells
promote

cell
proliferation and survival


Cytokines and chemo tactic
factors
produced by inflammatory
cells and other stroma promote cell
migration and invasion


Proteases

produced by the mE
break down basement
membrane
, altering the architecture
of tissue structures and
migration/invasion of tumor cells

Growth and
survival factors

Cytokines and
chemo tactic
factors

ECM

Proteases

Factors produced in cells within the
tumor microenvironment can alter
aspects of tumor cell behavior

Ca stem cell

Tumors are “organs” composed of
many interdependent cell types

New Challenges




New Disciplines






New Approaches

Cancer

Math

Engineering

Biology

Physics

Modeling,
informatics, ect

Basic, experimental,
clinical

Design, control
theory, application

Instrumentation, force,
thermodynamics, ect.

Understand, Prevent,
Detect, Treat

Mathematical Cancer Modeling

Black Box
-

no a
-
priori
information

Outputs

Signaling networks

Cellular Phenotypes

Drug Response

Tumor growth

Disease outcome

Inputs

Mutation
Status

Expression array

Protein Levels

Clinical
Parameters

Environmental
factors

Drug levels


White Box
-

all information known

Can Cancer be modeled
in silico
?


Complexity


Task vs. tools


Scale


Molecular
-

cellular
-

organismal systems


Input?
-

Output?


Types of data
-
quantitative, qualitative, amount


Read out


Model what?


Molecular interaction


Cellular processes

Cancer Modeling Spectrum


c
n
f
c
D
t
c
c










2
10010101001010
10

In vitro

Cell

Lines

Mammalian

Primates

Mathematical/Computational

“All models are wrong, but some are useful.”

George E. P. Box

Challenges and Promises of
in silico

Cancer
Modeling

Challenge


Diverse Data


“Tsunami” of Data


Diversity of Model scales and
outputs


Biology is Complex


Many unknowns


Models are wrong


Models are still wrong


Experimentation is expensive


Care is Personal

Promise


Contextual Data Integration


Scalable Infrastructure


Modularity of mathematics
and engineering


Models are Simple


Models make Predictions


Models can be tested


Models can be changed


Modeling is cheap


So are Models

Scaling Challenges

Modeling

Size

Population/
environment

Organism

Tissue/
organ

Micro
-
environment

Cell
(components)

Interactions

Molecules

Atoms

Statistical
mining

Bayesian
networks

Boolean
Models

Markov
chains

Differential
Equations

Time

Evolutionary

Generational

Lifespan

Tumor
development

Cell

Signaling

Reactions

Molecular/
conformationa
l




Cancer

Initiation

Micro
-
environment

Progression

Metastasis

Treatment

Regression

Recurrence



Robustness Control Theory


Fundamental organising principle of evolving dynamic systems


Allows essential functions despite perturbations


A system
-
level phenomenon, the individual components of a system may or may
not be robust themselves.


Highly Optimised Tolerance (HOT)
-
biological systems that have evolved to be
robust to run
-
of
-
the
-
mill perturbations are extremely sensitive to rarely
encountered perturbations. (John Doyle)


“robust yet fragile”, there is a trade
-
off between robustness, fragility, performance,
and resource demands.


Robustness may be regarded as a conserved quantity. This 'conservation of
robustness‘


Creating Robustness


Homogeneous redundancy is where multiple copies of components are used to support
the main system with many identical standby systems. This is rare in biological systems,
but common in engineered systems. Such a design is susceptible to common
-
mode
failure.


Heterogeneous redundancy is where biological systems possess genetic variability and
so are not vulnerable to the same insult. Archetictual involves networks and feedback

J. M. Carlson and J. Doyle,
PNAS
, 2002.

J. M. Carlson and J. Doyle,, 1999.



Normal
Cell

Cancer

“cured”

Fragility

Fragility

Robustness

Robustness

Cancer Evolution Theory

Mutation
environment

Treatment

Adapted from H. Kitano, J. Doyle

“Cell wants to maintain its current state”


Robustness
-
allows function in the face of uncertainty


Fragility
-

predicts catastrophic failure

Biology and Physics beyond Biophysics


Historically Physics has assisted biology with enabling
technologies


Electrophysiology, structural biology, membrane
mechanics


Go beyond this to theoretical approaches to explain
and test biological principals and phenomena


Avoid the “unifying” theory approach


"Chemistry is hard physics, and biology is hard
chemistry."

A Physicist Insight into Biology

Erwin Schrödinger


1944 published a treatise based on a series of lecture
entitled “What is Life?


living system exports entropy in order to maintain its
own entropy ,”negative entropy”, proposed the concept
of a complex molecule with code for heredity


Both Watson and Crick give credit to Schrödinger for
insight into their discovery of the nature of the DNA, the
molecule of life


But cautioned to classical physicists that “it is not that
the subject (biology) was simple enough to be explained
by mathematics, but rather that it was much too involved
to be fully accessible to mathematics”




The “S” Curve of Science

From Qualitative to Quantitative

Empirical

Observations

Data
explosion

Rise of Partial Theories

on complex systems

Search for Unifying

Theories


Growing Understanding

Of Subsystems

Adapted from E. Zerhouni

Systems Biology

Computational
Modeling

Data & Information
-

Clinical, Biological,
Epidemiological

Discovery and Knowledge
-

Basic and translational

Integrative

Cancer

Biology

(ICBP)

The ICBP Approach

The Iterative Nature of Systems Biology and
Modeling

Biological and Clinical
knowledge and data

Model Generation
and/or refinement

Prediction or
hypothesis

Experimental Design

Clinical or Basic
Experimental
Testing

Results and
Analysis

Hypothesis Testing
and Knowledge
Generation


Develop an integrative approach to the understanding of
cancer through the development of multidisciplinary
research teams



Create predictive
in
-
silico models to aid with the
understanding and management of the disease


Integrate and explore the multi
-
dimensionality of large
“omic” datasets as well as quantitative and descriptive
data.


Enrich the community and the developing field through
share resources and a vibrant educational/ outreach
effort.

Goals of the ICBP

Disiplines of ICBP Investigators
11%
28%
9%
19%
20%
13%
Medicine
Biology
Engineering
Math/Physics
Computer Science/ Bioinformatics
Chemistry
Disciplines of ICBP Investigators

Integrative Cancer Biology Program

Huang

(OSU)
-

epigenetics, gene silencing

Golub

(DFCI)
-

kinase, signaling, high throughput biology

Kinsella

(CWRU/UHC)
-

DNA repair, drug/radiation


effects, therapy

Hynes
(MIT)


signaling, mouse models, mitogenesis,

DNA repair, progression

Nevins

(Duke)


signaling networks, cell fate, proliferation


mouse models

Gray

(LBNL)


signaling, progression,


microenvironment, targeted therapies

Quaranta

(Vanderbilt)
-

invasion, metastasis,


angiogenesis, microenvironment

Deisboeck

(MGH)
-

angiogenesis, invasion


3D tumor modeling, repository

Cancer Cell

Initiation

Progression


Metastasis

Plevritis
(Stanford)


progression, lymphoma,


gene expression, clinical data


Lauffenburger,

Recent ICBP Activities


Development of validated siRNA library of cancer genes
http://cgap.nci.nih.gov/RNAi/shRNAValidation



ICBP 50 Cell lines


Over 35 computational models developed


AACR symposia


2009 ASCO symposium planned


Summer training program in integrative cancer biology


Post
-
doc exchange program


Multiple individual center workshops


Joint Cross discipline PI meetings


Cancer Modeling (MMHCC)/ TMEN


NIGMS Systems biology centers


Team building within and across centers


Data integration workshop


Recent meetings


November 2007, Washington D.C. ICBP mtg


May 2008, Columbus, OH ICBP mtg


EU
-
US workshop on Cancer Systems Biology

ICBP Centers
-

“Connecting Information in
Cancer Research”

Joe Gray (LBL) :
genomics
-
to
-
disease

Doug Lauffenburger (MIT):
molecules
-
to
-
cell function

Vito Quaranta:
(Vanderbilt)
-

cells
-
to
-
tissue

Tim Huang (OSU):
Cell
-
cell, genome

genome
-
Epigentic
influence across the
microenvironment

http://icbp.nci.nih.gov/

ICBP: Centers for Cancer Systems Biology
(CCSB)


Assemble interdisciplinary research teams focused on
integrating cancer biology with physical sciences.


Each Center will be required to have a multi
-
discipline team of investigators.


Develop and experimentally test predictive computational
models in the areas of basic and/or translational research


Develop educational, training and outreach programs at all
levels

Each Center will use systems approaches to
develop predictive computational modeling
technology and apply it experimentally to the
understanding of cancer:

Centers for Cancer Systems Biology (CCSB)


Proposed Requirements and NCI Review Criteria


Cancer biologist and physical scientists


Evidence of inter
-
disciplinary

research team


Strong cancer relevancy


Cancer Systems Biology Approach


Strong Mathematical modeling


Evidence of Educational programs


Institutional commitment to center


Mechanism


U54 Centers Grant


Cooperative Agreement


Budget


8
-
10 Centers


$2
-
3M per year per center

Collaborative Research in Integrative Cancer
Biology and the Tumor Microenvironment (U01)


This FOA is designed to: a) facilitate new projects in
integrative cancer biology and/or tumor
microenvironment research; and b) extend current
research conducted in the ICBP and TMEN programs
through collaborations with a broader research
community.


Research projects proposed in response to this FOA
should be collaborative and aligned with the missions of
the ICBP or TMEN


“adjunct” status in center programs


No set aside, but NCI directed review

Physical Sciences
-
Oncology Centers


To catalyze a fundamental level of understanding of the physical
and chemical forces that shape and govern the emergence and
behavior of cancer at all levels. These coordinated trans
-
disciplinary teams will develop and test innovative cancer
-
focused
hypothesis and create new fields of study based on the physical
laws and principles that operate in the biological spaces that are
critical to understanding and controlling cancer.


Understanding the Physics of Cancer


Exploring and Understanding Evolutionary Theory and Evolutionary
Processes of Cancer from a Physics Perspective


Understanding the Coding, Decoding and Transfer of Information
Transfer in Cancer


“De
-
Convoluting the Complexity of Cancer


Length Scale


In Reference to Size (Ranging from 1 nm


1 mm)

DNA

Protein

Cellular Components

Cells

Tumor

Evolution and Evolutionary

De
-
convoluting Complexity

Coding/Decoding/Transfer

Of Information

Physics (Physical Laws and
Principles)

Investigators tend to work in ‘length
-
scale’ silos. In order to have a
clearer picture of cancer, these
barriers need to be broken.

Summary of Theme Areas



2
-
4 Theme Projects

Administrative Core

Proposed Center Requirements

Overarching conceptual physical sciences
-

cancer theme/approach

Physical scientist PI with basic/clinical cancer researcher co
-
PI(s)

Transdisciplinary teams


and team environment

Adopt 2
-
4 synergistic theme projects (
i.e.,

complexity, coding, decoding,
transferring information, evolution/evolutionary theory, physical science
principles/laws)

Shared Research Capabilities collaboratively linked through multiple centers
including training, education, and outreach

External advisory board provides scientific input to the program

Proposed Center Structure


Shared Resources
and Capabilities

Governance Committee

Center 1

1
-
3 Institutions

Center 2

1
-
3 Institutions


2
-
4 Theme Projects

Administrative Core

Center 3

1
-
3 Institutions


2
-
4 Theme Projects

Administrative Core

Center N

1
-
3 Institutions


2
-
4 Theme Projects

Administrative Core

Science Focus
Group

Shared Resources
and Capabilities

Shared Resources
and Capabilities

Shared Resources
and Capabilities

Proposed Mechanism and Budget

Mechanism:

U54 (Cooperative Agreement Specialized Center)


Facilitate and enable the convergence of the physical sciences with
cancer biology to understand cancer at a fundamental level


and create
new opportunities for intervention and control


Need for several disciplines and teams to form a network and
governance structure


Active coordination by program managers trained in physical and cancer
sciences
-

and input from scientific focus groups


Accelerate leverage gained from resource sharing


Provide linkages to key NCI programs and resources


Budget:

4
-
6 Centers

$3.0
-
3.5M per year per center

Annual budget


$15
-
21M, Total budget


$75
-
105M

Cancer Biomedical Informatics Grid (caBIG)


Common, widely distributed infrastructure permits cancer research
community to focus on innovation


Shared vocabulary, data elements, data models facilitate
information exchange


Collection of interoperable applications developed to common
standard


Raw published cancer research data is available for mining and
integration

caBIG In Silico Research Centers (ISRCE)

-
New NCI CaBIG RFP
-


G
enerate
publishable novel research findings through integrating, analyzing
and/or mining the data accessible through caGrid

and other publicly available
data sources;


Identify
novel processes (including workflows) and tools that exploit these data
resources by supporting the analysis of data generated by early phase clinical
and basic research studies;


Develop
and make publicly available new data analysis, aggregation and mining
algorithms capable of working with disparate, potentially distributed data
resources;


Encourage
availability of additional data resources and analytic services that will
be useful for biomedical research; and


Assess
the capacity of current publicly available data resources to support
cancer research through:


discovery of non
-
obvious relationships


determination of missing key attributes and links that could enhance analysis.


aimed primarily at bioinformaticians, computational biologists, and statisticians who are
asking key biologic and medical research questions that can be addressed through data
mining, analysis and the integration of the wide array of disparate datasets
.


RFP Applications due February 5, 2009



NIH Funding Opportunities in Systems and
Computational Biology


Various scale R21 to “center”


Individual projects to Team Science


Directed Research and “Investigator Initiated” Research


Collaborations are good


Communicate with scientific program staff first


Many published specific opportunities
http://grants.nih.gov

National Centers for Biomedical Computing
(NCBC): an NIH Roadmap Program

The NIH NCBC will be devoted to all facets of biomedical computing,
from basic research in computational science to providing the tools
and resources that biomedical and behavioral researchers need to
do their work. Currently 7 centers funded, Planned re
-
issuance
2009


Collaborations with National Centers for Biomedical Computing
(R01)


PAR
-
07
-
249
-

Released December 20, 2006


Exploratory Collaborations with National Centers for Biomedical
Computing (R21)


PAR
-
07
-
250
-

Released December 20, 2006


BISTI at the NIH


The
Biomedical Information Science and Technology Initiative

is a
consortium of representatives from each of the NIH institutes and centers.


The mission of BISTI is to make optimal use of computer science and technology
to address problems in biology and medicine by fostering new basic
understandings, collaborations, and transdisciplinary initiatives between the
computational and biomedical sciences.



Innovations in Biomedical Computational Science and Technology (R01)


PAR
-
07
-
344

-

Released March 29, 2007.


Exploratory Innovations in Biomedical Computational Science and Technology (R21)


PAR
-
06
-
411

-

Released May 15, 2006.


Continued Development and Maintenance of Software (R01)


PAR
-
08
-
010

-

Released
October 18, 2007


Innovations in Biomedical Computational Science and Technology Initiative (SBIR
[R43/R44])


PAR
-
07
-
160
-

Released December 12, 2006.


Innovations in Biomedical Computational Science and Technology Initiative (STTR
[R41/R42])


PAR
-
07
-
161
-

Released December 12, 2006


Data Ontologies for Biomedical Research (R01)


PAR
-
07
-
425

-

Released August 3, 2007


Sharing Data and Tools: Federation using the BIRN and caBIG Infrastructures (R01)


PAR
-
07
-
426
-

Released August 3, 2007




Interagency Modeling and Analysis Group (IMAG)



Interagency Modeling and Analysis Group (IMAG) is an internal
working group, comprised of program staff from nine Institutes of
the National Institutes of Health (NIH) and three directorates of the
National Science Foundation (NSF). IMAG now represents 17 NIH
components, four NSF directorates, DOE, DOD, NASA), USDA,
and the VA


Currently 24 R01 grants awarded through multiple agencies


Predictive Multiscale Models of the Physiome in Health and
Disease (R01)


PAR
-
08
-
023


re
-
Released November 5, 2007.

Future NCI Programs Linkages: Continuum

ICBP

*PS
-
OC

Centers

Nano Alliance

IMAT



R01

TCGA

CPTAC

MMHCC

caBIG/NCBI

caHUB

Other

New Knowledge/Physical Sciences Cancer Systems Nano
-
Based High Dimensional
Interventions



Future Challenges for Investigators in Integrative
Biology


Team research vs. Individual research


Identification of issue


Experimental design


Coordination and evolution of the science


Recognition at all levels


Scientific career path


Focus of research: question vs. approach


Establishment of independence


Choice of new environments


Very dynamic and exciting time in biomedical research


Institutions have to catch up to individuals and science


Future Directions of Cancer
Research


Integration of Disciplines


Leverage efforts and strengths


Address technological needs



Integration of Institutions


Funding (private, governmental, international)


Training



Integration of Thinking


Individual vs. Team Science balance


New approaches and Ideas



Thank you