UBM Research Project Summaries
The project pages will open with Adobe¨ Acrobat¨ Reader¨
Click Here for Adobe¨ Acrobat¨ Reader¨ Ñ free software for viewing and printing PDF files.

Evaluating Parallel Approaches for Evolutionary Trees (click for a larger view of the image).

Research Mentors: Kevin Omland (Biological Sciences), Matthias K. Gobbert (Mathematics/Statistics)

One of Charles Darwin's main visions was reconstructing a "Tree of Life" showing evolutionary relationships among all species on the planet. Recently, evolutionary biologists have been using new genomic data to reconstruct many parets of the tree of life. However, determining relationships among closely related species remains one of the greatest challenges, especially because closely related species have evolved very few species-specific differences in theirDNA sequences. Our research group is using multiple gene regions from different parts of the genome to reconstruct these evolutionary family trees - phylogenies. Many new "species tree" algorithms have been developed to analyze such data for recently diverged species. We work on a genus of birds that has thirty species, plus multiple subspecies. Some of the new programs that we are using (e.g., BEST) require weeks to run on computer clusters for just five or ten species. To scale up our analyses to fifty or more species and subspecies, we need much more computationally efficient approaches, including parallelization. A newer perogram (*BEAST), which may allow parallel analyses, has just been released.

Evolution of Stochastic Gene Transcription Networks
Research mentors: Ivan Erill (Biological Sciences), Muruhan Rathinam (Mathematics/Statistics)

Transcriptional regulation is a fundamental step in the gene expression cascade. Gene transcription is regulated by the coordinated activity of transcription factors, which often organize themselves into small networks, or motifs, capable of generating complex temporal patterns of gene expression. Transcription is a noisy process and capturing the full repertoire of expression profiles produced by different network motifs requires hybrid stochastic and deterministic simulation methods. Experimental data and genomic surveys suggest that living beings tend to exploit shallow network motifs with few components. Here we propose to investigate the evolutionary rationale for this design strategy by evolving arbitrarily large networks in the context of a genetic programming framework based on a hybrid transcription simulator.

 

The roleof TRPM5 in olfactory sensory neurons: A model for understanding membrane ion channels in sensory signaling
Research mentors: Weihong Lin (Biological Sciences), and Jonathan Bell (Mathematics and Statistics)


The mammalian olfactory system is capable of detecting thousands of airborne chemicals.  Signal transduction, i.e. a process of transferring chemical energy into electrical impulses, is mediated by a G-protein coupled cAMP signaling cascade in which the binding of odorants to odor receptors results in activation of the cyclic nucleotide gated (CNG) channels. Recently another signaling ion channel, TRPM5 was identified in some olfactory sensory neurons with unknown function.  UBM students will use electrophysiological recordings and mathematical modeling to determine the dynamic contribution of these ion channels. The picture shows an olfactory sensory neuron labeled with olfactory marker protein and TRPM5.

 

Modeling a Cellular Response to a Gradient

Research mentors: Michelle Starz-Gaiano (Biological Sciences), and Brad Peercy (Mathematics and Statistics)

This project investigated regulation of the conserved JAK/STAT signaling pathway, which is essential for cell migration.  STAT activation is initiated by diffusible molecules radiating from a localized source. To analyze how cells respond to a gradient of signaling, we examined the ovary of Drosophila.  High levels of STAT activation cause a cluster of cells to migrate, while nearby cells with lower activity remain stationary. A balance of opposing transcription factors, SLBO and APT, determines the correct number of motile cells. To identify the minimal requirements for how cells can resolve a graded signal into a step-wise activation, we have developed a mechanistic mathematical model based on known/probable molecular interactions. 

 

Modeling Collective Cell Migration

Research mentors: Michelle Starz-Gaiano (Biological Sciences), and Brad Peercy (Mathematics and Statistics)

We are developing a 2-dimensional simulation of the acquisition of cell movement by epithelial cells, then the subsequent migration of a cluster of cells and re-closure of the epithelium.  In this model we designate which cells are migratory and determine the various intrinsic and extrinsic forces that act upon them and move them accordingly. This allows us to track the migratory cluster as it moves towards the oocyte. Our initial testing with this model indicates that the simplest proposed biological mechanisms for the movement are sufficient to allow the cluster to successfully traverse the egg chamber and undergo some amount of rotation, as seen in vivo.  This work may be expanded into three dimensions. 

 

Modeling Response to Non-Uniform Spatial Distribution of a Morphogen

Research mentors: Michelle Starz-Gaiano (Biological Sciences), and Brad Peercy (Mathematics and Statistics)

The distribution of an activating signal diffusing from a point source is impacted by the environment through which it is released.  We are examining the impact of the intercellular space created by adjacent tissues on the ability for the extracellular signal to activate a cell fate decision heterogeneously.  We will study this effect using confocal imaging, genetic manipulation, and spatial models with appropriate geometry.

   

 

The JAK/STAT signaling pathway in a simple epithelium: A model for understanding molecular interpretation of spatial gradients

Research mentors: Michelle Starz-Gaiano (Biological Sciences), and Brad Peercy (Mathematics and Statistics)

 

The JAK/STAT signaling pathway plays a critical role in stem cells, immune function, and the progression of some cancers.  We are studying JAK/STAT signaling in the somatic epithelial layer of Drosophila ovaries, where graded pathway activation is refined to instruct certain cells to become motile (those labeled in green in the egg chamber picture) while other cells remain behind in the epithelium (labeled in red).  We are developing mathematical models to determine critical parameters in this signaling system.  Our goal is to understand the minimal biological components that can convert a gradient of information into binary activation of a molecular pathway. UBM students working on this project will use a combination of genetic, cell biological, and mathematical approaches to address this aim. The picture shows an egg chamber in which border cells, marked in green have migrated away from their starting site, the anterior epitheliumk, shown on the left by the red cells.

     

The mouse light response of intrinsically photosensitive retinal ganglion cells: A model for understanding the second messenger pathway
 Research mentors: Phyllis R. Robinson (Biological Sciences), and
Kathleen A. Hoffman (Mathematics and Statistics)

Modeling of light entrainment of the mammalian circadian clock requires input exclusively from the retina. This photo-response is unique and also expresses melanopsin, a novel vertebrate opsin, which is necessary for initiating the light response in these cells.  Among all known vertebrate opsins, melanopsin is unique. These light sensitive ganglions cells were just discovered 10 years ago and the biochemistry underlying the light response remains to be definitively elucidated. The goal of our joint research project is to model the light dependent depolarization. The drawing depicts the crystallized molecular structure of melanopsin.

 

 

Simulating the evolution of transcription factors and their binding sites
Research mentors: Ivan Erill (Biological Sciences), Matthias Gobbert (Mathematics/Statistics)

Transcription factors govern the regulation of genes by binding to specific locations in the promoter region of genes, enhancing or repressing their transcription by the RNA-polymerase holoenzyme. The mechanisms by which transcription factors recognize their cognate binding sites in DNA remain obscure and many questions about the evolution of these genetic elements remain unaddressed. Here we propose the use of UMBC parallel computing facilities to simulate the evolution of virtual transcription factors operating on synthetic genomes in order to analyze the evolutionary constraints operating on transcription factors and their binding sites. A combination of experimental data and combinatorial approaches to simulate evolutionary processes can shed light on the nature and evolution of transcription factor-finding motifs like the one depicted in this picture.