Seminar Schedule Spring 1998
- 14 January
Wilson Radding, Rice University
"Using genetic algorithms to actually model genetics:
a proposed simulation of the evolution of HIV-1 envelope
glycoprotein in a single patient"
ABSTRACT:
HIV wages a war with the infected person's immune system.
The front line of the war is fought
by CD4+ T cells, which are killed and replaced by the billions each
day during the asymptomatic phase of the disease. This phase can last
10 to 15 years (in rare cases indefinitely). Within a given patient
during this time the virus evolves, sometimes to the point where variable
regions of the envelope glycoprotein retain few of their original amino
acids. All the circumstantial evidence indicates that this evolution
represents escape from immune surveillance, promoted by the incredible
lack of fidelity of the HIV reverse transcriptase, which in in
vitro assays introduces errors as often as every 500-600 bases. Models
of this viral evolution have relied on standard evolutionary differential
equation predictions. In a general way these models can predict the
ascent and descent of new substrains within a patient, but they have no
way of connecting the process to specific events in a viral life cycle.
We are at the beginning of an attempt to model the war between the virus
and the immune system by designing two interacting finite automata, one
representing the viral env gene and its products, gp160 plus parts of nef,
tat, rev and vpu, and one representing the relevant parts of the immune
system, B cells and T cells producing immune responses to the portions of
gp160, gp120 and the exterior segment of gp41, which are readily
available.
-
28 January
Ralph Zinner, M.D. Anderson Cancer Center
"DNA chip technology"
-
11 February
Luciano Bono, University of Houston
"Modeling central pattern generators for animal gaits using symmetric ODEs"
-
25 February
Shane Pankratz, Rice University
"Branching Processes and Linkage Disequilibrium:
Estimating the Recombination Coefficient"
ABSTRACT: Classical methods for the genetic mapping
of disease genes suffer from limits in resolution.
Much work is being done to derive alternative procedures.
This talk discusses the role that branching processes can
play in genetic mapping via linkage disequilibrium.
-
11 March
Steven J. Cox, Rice University
"Inverse Problems in Biomechanics: Recovering the constitutive
law from observed deformation under a prescribed load"
-
25-26 March
Janet Siefert, Rice University and University of Houston
"Workshop on Molecular Evolution"
25 March, 1 - 3:30 PM
Molecular Systematics
Introduction and Principles of phylogenetic inference
Applications, examples, and problems
Practical application, protocols, and available tools
26 March, 1 - 3:30 PM
Molecular evolution: getting results from systematics
Phylogenetic mapping
Whole genome sequencing and analysis
Application to bacterial evolution
Phylogeny of ribozymes
-
8 April
B. Montgomery Pettit, University of Houston
"Microfolding and Phase Stability in Peptide Solutions:
Integral Equation Method in Computational Biology"
-
29 April
James R. Thompson, Rice University
"SIMEST and Clinical Trials"
ABSTRACT: There is a practical inability
to get from stochastic process axioms of cancer progression
to models at the level of aggregate clinical data.
Consequently, there has been a tendency to use survival
data on linear models for clinical
trial data as a means of making sense of cancer progression.
After over fiftyyears and many billions of dollars expended,
this clunky Mengelian approach has proved ineffective.
In the "War on Cancer". The score
is Cancer 35, Visitors 2. We show how it is possible to use
high speed computing to get from the stochastic process model
to the aggregate model in a way that the underlying parameters
of the model can be estimated using clinical data. The SIMEST
approach also has the potential of taking the field
of stochastic processes from arcane irrelevancy to center stage
in applied modeling in many fields from economics to biometry.
Return to previous page