Transcript advisory_council - University of Notre Dame

Computational Biology and
Bioinformatics at Notre Dame
Prof. Jesús A. Izaguirre
Department of Computer Science and Engineering
Overview

Systems approach to computational biology and bioinformatics





Research in CSE




Biomedical engineering
Biocomplexity simulations
Computational biology and biochemistry
Challenges




Integrate multiple levels of information
Distinguish noise from signal: analysis of large data sets
Mathematical modeling of biological complexity
Computer assisted analysis of integrated data
Right biological problem and team
Obstacles to interdisciplinary research (publications, learning curve)
Computational resources
Opportunities






Indiana Biocomplexity Consortium
Interdisciplinary Center for the Study of Biocomplexity
Collaborative frameworks
Systems biology approach
Approximate multi-scale, multi-level models
Nanoscale genomics
Background on computational
biology and bioinformatics I
How to integrate data from many levels?
1.
Genomic sequencing data
2. Proteomics data (global analysis of
proteins)
3. Gene expression data (DNA
microarrays)
4. Protein-protein, metabolic networks
5. Genetic regulatory networks
6. Morphogenesis: cellular and genetic
Background on computational
biology and bioinformatics II
How to distinguish noise from signal?
1.
Growth in number of base pair
sequences has surpassed Moore’s law
about the number of transistors in a
chip
2. Dr. Leroy Hood predicts that in 10
years, using nanotechnology to analyze
single cells and molecules, one will
sequence human genome in 1 day
Background on computational
biology and bioinformatics III
Mathematical modeling of biological
complexity
1.
Multiscale in nature
2. Several qualitative models
1.
2.
3.
3.
Rule based – intuitive for biologist
Continuum based – PDE – can be linked to
quantitative experiments
Discrete – Monte Carlo or Molecular
Dynamics
Need to relate more to digital code
Background on computational
biology and bioinformatics IV
Computer assisted analysis and
visualization of integrated data
1.
Multi-tool in nature
2. Geographically distributed, collection
of cooperating tools
3. Visualization and analysis should
integrate both simulation and
experimental data, to facilitate cross
validation and comparisons
4. Dissemination via web services
Current Research at ND CSE I
Danny Chen and students are
developing algorithms for
radiosurgery treatment
planning
•Find a set of beams to destroy
a tumor without harming
surrounding healthy tissues.
•Considerations:
–Desired dose
–Constraints of the device
delivering the radiation
–Treatment time
•Collaboration with Univ. of
Maryland radiologists
Current Research at ND CSE II
Greg Madey and Patricia Maurice (CE) are
developing a collaboratory for biocomplexity
simulation of environmental chemistry
•Try to model the life cycle of pollutants in a soil –
swamp ecosystem
• Validated through experiments
• Web based interface using Java and SWARM
simulation package
• Simulation results are stored in an Oracle
database (“E-science”)
• Multi-institutional collaboration funded through
NSF Biocomplexity grant
Research in Izaguirre’s group I
Jesús Izaguirre and collaborators are
working on:
Computer-aided drug design, human
genome proteomics, and
understanding of morphogenesis
by:


Enabling simulations of
biomolecules (Proteins, DNA,
etc.)
Permitting simulations of cells
and organ growth
CHALLENGES:
• Large systems--millions of particles
• Long time scales--billions of time
steps
=weeks and months of simulations in
supercomputers with hundreds of
nodes!
Research in Izaguirre’s group II
We have released CompuCell,
a multi-model software
for simulation of
morphogenesis
1.
Models interaction of
genetic regulatory system
with cellular dynamics
2. Uses knowledge-base,
stochastic, discrete and
continuum model
3. NSF Biocomplexity grant
to model chicken limb
development
Research in Izaguirre’s group III
Molecular dynamics of biological molecules:
Faster algorithms, up to an order of magnitude faster
for molecular dynamics and sampling of
conformational space of proteins
Parallel program ProtoMol, a software framework for
molecular dynamics and related-applications, which
is open source (http://www.nd.edu/~lcls/protomol)
IN PROGRESS:
• MULTISCALE approximate methods in biology and MEMS (with
Paolucci)
• Applications in real problems:
• Brian Baker in Chemistry (immune system)
•Ruhong Zhou at IBM T. J. Watson (BLUE GENE project)
• Grant
application to NSF Nanoscale. Army in
preparation
Opportunities I
Indiana Biocomplexity Consortium
Simulation Request
OBJECTIVE: Make ProtoMol and
CompuCell into web services



Web service for molecular and
cellular simulations
Component that provides data and
simulation capabilities through the
web
NIH Center of Excellence proposal
(Indiana University and Notre Dame)
Results via XML
CompuCell Biocomplexity
Computing Environment
Server
APPROACH:
• Integrated tools for simulation,
visualization and analysis
• Use distributed grids
• Disseminate using web services
Opportunities II
Research areas of interest
Collaboratory frameworks:
recommender systems, assistants, not
just algorithms
 Data mining in high performance
environment: clustering, pattern
recognition, distributed databases
 Modeling: constrained optimization,
stochastic techniques, geometry
 Simulation: qualitative/quantitative
models

References
I would like to thank the following-NSF Biocomplexity grant IBN-0083653
NSF CAREER award ACI-0135195
Department of Computer Science and Engineering,
Univ. of Notre Dame










http://conferences.computer.org/bioinformatics
http://smi-web.stanford.edu/people/altman
http://geneticcircuits.ucsd.edu
http://www.systemsbiology.org
http://www.geneontology.org
http://www.cytoscape.org
http://www.genomatica.com
http://www.nd.edu/~lcls/compucell
http://www.nd.edu/~lcls/protomol
http://www.nd.edu/~xwu/uiowa_files/v3_document.htm