Fiscal Year 2001 Awards

Division of Biological Infrastructure

Postdoctoral Research Fellowships in Biological Informatics



Fellow's Name Host Institution Research Area/Training Plan NSF Award #
Title of Research and Training Plan
Abstract


Ilya Balabin University of Illinois, Urbana-Champaign Biophysics 0107594
Computer simulations of biological molecules: obtaining unique information on their structure and function from STM images
Novel methods and tools for computer simulations of scanning tunneling microscopy (STM) images of biological molecules are being developed. These methods are being used for obtaining unique information about structure, dynamics, and function of proteins, lipids, and nucleic acids from their STM images. The methods will also facilitate tailoring STM experiments.


Brian Beckage University of Tennessee Population Biology 0107553
Modeling the effects of hurricane and fire disturbances on the population dynamics of south Florida slash pine
Fires and hurricanes are frequent disturbances affecting community structure and dynamics of pine forests in the southeastern US. This research is developing spatially explicit, individual-based models of pine population dynamics, parameterized from empirical data that include hurricane and fire disturbances. The models are to be used to predict how pine populations will respond to hypothetical disturbance regimes associated with global change.


Nicolas Buchler Univ California, San Diego Developmental Mechanisms 0120292
From genome to gene expression: mapping and understanding genetic networks in Dictyostelium development
Upon starvation, Dictyostelium amoeba communicate, sort, and adhere to each other to form ``tissues''. This transformation, resulting in a resistant spore, utilizes all the mechanisms crucial for normal development of multicellular organisms. Promoters, repressors, and the architecture of genetic networks involved in development in Dictyostelium are being explored using both genomic and gene expression data and mathematical modeling.


Daniel Butts Harvard Medical School Neurobiology 0107581
How neural activity instructs the system-wide modifications in brain structure that occur during development of the visual system
Spontaneously generated "retinal waves" are required for early visual system development. This research uses a combined theoretical and experimental approach to determine (1) how physiologically relevant neural input causes appropriate modifications at individual retinogeniculate synapses, and (2) how these retinal-wave-driven modifications result in system-wide competition that drives retinogeniculate development.


Duncan Callaway University of California, Davis Ecology 0107571
Dynamic species interactions, uncertainty, and the maintenance of ecosystem robustness
The interplay between community complexity and stability is the source of one of the classic debates in ecology. Heterogeneous spatial and temporal species distributions and trophic interactions, though central to the debate, have been studied little. To address these issues this research develops new computational methods to parameterize, simulate, and analyze a multi-species, multi-patch model.


Will Fischer University of Texas, Austin Systematic Biology 0107569
Phylogenetic analysis using ribosomal RNA structure
Ribosomal RNA genes are widely used in phylogenetics and have shaped current views of the history of life. Traditional DNA sequence-based phylogenetic methods, however, fail to reliably place highly divergent sequences. This research develops probabilistic models for phylogenetics that are based on rRNA structure.


William A. Frankino University of Leiden Evol & Ecol Physiology 0107577
The role of epigenetic signaling in the evolution of trait allometries: integrating development with morphometrics
This series of studies is determining the proximate mechanisms by which morphological trait allometries evolve in insects, specifically the wings and bodies of butterflies. Changes in the shapes of, and allometries among, morphological traits are being produced by artificial selection. The response to selection is being quantified using statistical and morphometric methods developed by the Fellow.


Audrey Gasch Lawrence Berkeley National Lab Eukaryotic Genetics 0107562
Development of alternative clustering methods for genomic expression analysis and characterization of coregulated genes in yeast
This project is developing fuzzy k-means clustering for analysis of genomic expression data, with the goal of identifying coregulated yeast genes that are not identified by traditional methods. After identifying sets of coregulated genes, multiple algorithms will be applied to identify sequence motifs that contribute to the regulation of expression.


Jessica L. Green Macquarie University
University of California, Davis
Systematic Biology 0107555
Estimating global microbial diversity: a bioinformatics approach
Scaling law, probability, and combinatorial theory are being used to quantify biodiversity patterns across a wide range of spatial scales and at different levels of phylogenetic resolution. Using computer intensive spatial statistical analyses and simulation modeling, hypotheses are being tested with a microbial genetic database two orders of magnitude larger than any previously reported (~73,000 16S rDNA clones collected over ~ 1 m2 to ~10,000 km2 in Australia, analyzed by RFLP and other sophisticated genetic techniques).


Susan Harrington Columbia University Biophysics 0107544

Using local symmetry to improve protein structure prediction

Finding minima of a potential energy function is a crucial part of many approaches to protein structure prediction. The goal of this research is to exploit a geometric relationship among the minimizing structures of a specific potential to efficiently find distinct minima. Transformations are being automated that take one minimizing structure to (or near) a different minimizing structure. Analogous relationships among homologues and the interplay of such transformations and domain classification systems are also being studied.


John Hutchinson Stanford University Animal Physiology 0107574
How dinosaur locomotion evolved: using 3D computer models to test biomechanical hypotheses
This research constructs realistic 3D computer models to visualize the physics of hindlimb muscle function in an alligator, chicken, and three extinct theropod dinosaurs. These models permit the testing of hypotheses like "Could Tyrannosaurus run?" and "How did the bipedal stance of birds evolve?" with new analytical rigor.


Louise Mead Oregon State University Evol & Ecol Physiology 0107583
A new approach to modeling the evolution of sexual isolation
This project explores the evolutionary dynamics of sexual isolation in salamanders and Drosophila. Theoretical models are being used to simulate evolution of sexual isolation given a phylogeny. Simulated data matrices will be compared to empirical data, providing the first tests of the roles of inheritance, drift, selection, and reinforcement in the evolution of sexual isolation across clades.


Christine Muirhead Harvard University Eukaryotic Genetics 0107596
Inferring the nature of quantitative genetic variation: methods for integrating single and multi-locus analyses

The goal of this project is to extend quantitative genetic theory to the study of individual genes, and then to use that theory to develop methods for analyzing sequence polymorphism data from quantitative trait loci (QTL). This theoretical framework allows the testing of hypotheses about the nature of genetic variation.


Eric Perreault Northwestern U Medical School Animal Physiology 0107553
Modeling Muscle Force Generation during Posture and Locomotion
This project involves developing models of muscle force generation. The hypothesis being tested is that, during functional conditions, whole muscle models can be constructed by summing responses of constituent motor units. Specific aims are to test this hypothesis experimentally and to use nonlinear system identification to generate a database of single motor unit models.


Ronald Peterson Univeristy of Pennsylvania Biochemistry 017595
Development of strategies for designing novel catalytic enzymes

This work includes an informational approach followed by experimental application of the results. The informatics portion includes parameterization of a whole protein energy function based on mutant to wild-type redesigns of proteins from various structural databases. Second, a comprehensive cataloging of specific interactions in natural cytochromes is being conducted, noting the conformational positioning of the reactive centers of these molecules. The strategies developed from these sections will be applied experimentally to design a novel synthetic cytochrome with heme binding and reactivity approaching that of its naturally occurring counterparts.


Ellen Pritham Marine Biological Laboratory Population Biology 0107590
Determining population structure and dynamics of Class II transposable element families in Arabidopsis thaliana, using a coalescence-based population genetic approach

Transposons are major constituents of most eukaryotic genomes that can dramatically alter genomic architecture, individual fitness, and species evolution. Little is known about population dynamics and behavior of transposons. In this research, the availability of complete genome sequence and biological informatics capability provides a novel opportunity to evaluate their impact and the forces that influence their movement.


George C. Stecker Univ Michigan Medical School Sensory Physiology 0107567
New techniques for analyzing the representation of auditory stimulus features by ensembles of cortical neurons
This project is concerned with the analysis of stimulus-related information contained in firing patterns of ensembles of 16 or more simultaneously-recorded neurons in auditory cortex. Statistical pattern recognition techniques are used to characterize relationships between the acoustic properties of sounds in space and information-bearing features of neural responses.


Christopher M. Summa Stanford University Biophysics 0107565
Ab-initio prediction of membrane protein structure

Membrane proteins which play important roles in cell function are not as well-studied as soluble proteins. This research develops automated or semi-automated computational methods for predicting the structures of integral membrane proteins using publicly available data and constructs an online database of structure predications. In doing so, it brings together information from protein folding theory, the human genome project, and mutagenesis experiments on membrane proteins in a unique way.


 
Dennis Wall Yale University Systematic Biology 017592
Exploring phylogenetic trends of gene sequence, protein form, and function to comprehend processes of molecular evolution

The photosynthetic gene rbcL and a large phylogeny are being used to link molecular selection linking gene sequence to protein form and function. Evolutionary trends are being used to reveal patterns of molecular selection that are otherwise invisible and are helping to generate viable theories about gene-protein relationships. The research is focused on developing generally applicable models of gene and protein evolution and providing informatic tools to question processes of molecular selection in other genes and systems.