My unofficial transcript of classes at Chico State reflects my interest in human health and disease, and fulfills the unit requirements for the Masters Degree in Biology.

I was the primary animal care technician for the Chico State Biology Department from December 2003 to May 2005. This involved basic care of rats and mice as well as injections, euthanization, and data collection for experiments.

I wanted to find a Masters Thesis project that was both relevant and meaningful, yet also do-able in two years. My initial interest was in the malaria mosquito Anopheles gambiae, and the mechanism of action of insect repellents. I found studies that showed that DEET was very toxic in animal models, and it's mechanism of action was unknown. John Carlson at Yale University had developed a DEET-insensitive Drosophila mutant, and when I inquired regarding their progress in isolating the gene(s) after their initial publication on this, they responded that they were maintaining the stock, were not doing anything with it, and that I could have samples.

I developed this project independently, and Jeff Bell was my advisor in the true sense of the word, in that he gave me advice when I faced problems and needed feedback. The experiment design, statistical analyses, as well as the theoretical basis of my hypothesis for the candidate gene was based on my original work and observations. This project included the lab work required for DNA extraction, PCR amplification, and sequencing reactions required to sequence a candidate region in both mutant and wild-type strains. I used wrote a data conversion script to load the recombination data into QTL Cartographer for quantitative trait mapping.

My poster presentation of this project won the highest honors in the Spring 2005 Biology Poster Seminar at Chico State, the Tom Rogers Outstanding Student Researcher Award, and the written thesis won the campus-wide Outstanding Masters Thesis, awarded in 2006.

My goal with this project is to create a web-based software application that automates the iterative BLAST searches and fragment assembly process that is required when searching for contiguous regions of DNA from the NCBI Trace Archive, an online database of single-read DNA fragments from shotgun cloning projects. Although there may be enough data in this database to assemble regions of the genome that are of interest to specific researchers, currently this must be done manually, or the researcher must wait until the entire genome is assembled.

I have implemented this functionality using Perl/CGI, BioPerl modules, and String::Approx. My prototype web interface allows researchers to search multiple organisms for a single sequence and receive results via email. With over 200 unique organisms in the Trace Archive, this tool could become an important resource for biologists.

T-RFLP (Terminal Restriction Fragment Length Polymorphism) is a way to characterize the diversity of microbial communities, and in this experiment we compared the microbial diversity of soil in contact with wild-type and mutant Arabodopsis plants, where the mutants constituitively expressed known plant defense genes. Our goal was to see if upregulated plant defense had an effect on the microbial community.

We PCR amplified the DNA corresponding to the 16S small-subunit ribosome of bacteria in our soil samples, and digested these samples with a single restriction enzyme, HhaI. We used an ABI 310 Sequencer to obtain T-RFLP raw data, and I wrote a perl program for data compilation and analysis. This script has many options and features, since we wanted to explore different ways of grouping and analyzing the dataset. This script generates similarity scores that can be input into Cluster.exe or averaged data that can be exported to Excel to generate electropherogram profiles that have been averaged across treatment groups.

One of the most highly studied examples of predator-prey relationships is on Isle Royale, Lake Michigan. Moose make up 90% of the diet of wolves, and field data has been collected since 1955. Previously researchers have examined this dataset to examine how spatial relationships effect population dynamics.

My goal was to see if the spatial relationships of Isle Royale are duplicated in an agent-based simulation, will the population response also mimic the field data. I modified an open source agent-based predator/prey simulation written in Java where neural networks are inherited (with mutations) by surviving individuals, and thus individuals within a species evolve behaviors based on the environmental conditions.

Although the outcome of my simulations were often similar to field data, there were also significant differences in some situations, thus I concluded that there are many factors underlying the success of predators on Isle Royale, and that simple models such as this one do not take into account complex interactions in the real world.