Student Research Symposium

Saturday, October 14th, 2017, 9am-12pm


Oral Presentations 9 a.m.-10:30 a.m., Richter Hall, Ahmanson Science Center

Poster Session  10:30 a.m. - 12:00 p.m.,  Ahmanson Science Center Lobby

View Past Symposiums

Saturday, October 14, 2017
9:00am - 10:30am
Oral Presentations

Location: Richter Hall, Ahmanson Science Center

Welcome remarks, Overview of the session, Presenation of awards, and Research Talks

Toward the Efficient Syntheses of DNA Dye Molecules: Investigating the Reaction Scope with Various Secondary Amines

Program: John Stauffer Research Fellowships in the Chemical Sciences
Faculty: Dr. Jason Kingsbury, Biology
Student: Emily Armbruster

Dye molecules are one of the tools used to study DNA.  These dyes attach to DNA strands and glow when exposed to specific wavelengths of light, allowing detection of DNA-rich areas in cells or electrophoresis gels.  The most common dye, ethidium bromide, is toxic, posing a risk to biologists as well as the natural environment.  Our study is part of a program to develop dimethylaminonaphthylpyridinium as a nontoxic and more versatile fluorophore.  This summer, a direct amination methodology was examined.  In the step investigated, 6-bromo-2-naphthol undergoes a formal nucleophilic aromatic substitution reaction with a secondary amine as the nucleophile.


Purpose: The purpose of this study was to examine how the electronic and steric character of various amines affect the reaction’s efficiency.


Methods: The three amines selected for this study, 4-piperidinemethanol, diisobutylamine, and azetidine, represent a range of steric and electronic characteristics.  Each amine was mixed with 6-bromo-2-naphthol under conditions reported in the literature. After work-up, column chromatography and recrystallization, melting point, 1H NMR spectroscopy and mass spectrometry were used to determine the purity and identity of the products.  The reactions’ relative efficiencies were measured by comparing each reaction’s percent yield.


Results: The reactions with 4-piperidinemethanol and diisobutylamine produced the desired pure products. The 4-piperidinemethanol reaction’s efficiency was twice as high with an average percent yield of 9.6% in comparison to the 4.4% yield of the reaction with diisobutylamine.  The azetidine reaction did not yield isolable product.


Conclusion: The diisobutylamine reaction’s lower efficiency is likely due to steric hindrance and lower water solubility. For the azetidine reaction, lack of isolable product may have been caused by the compound’s high-strain ring breaking under harshly basic conditions.  Now that two new potential substrates for fluorophore synthesis have been identified, reaction optimization can begin for 4-piperidinemethanol and diisobutylamine

Why do Squirrel Monkeys Urine Wash? A Field Study in Amazonian Brazil

Program: Jung Summer Research Fellowship for Biological Science
Faculty: Dr. Anita Stone, Biology
Student: Jacob Jasper

Many New World primates perform a behavior called urine washing (hereby UW), in which an individual will urinate into the palms of their hands and wipe the liquid on the soles of their feet. Though UW has been extensively studied in captive monkeys, the adaptive function of this behavior remains a mystery.The purpose of my study was to test four hypotheses on the function of UW by examining what social and environmental factors contribute to its frequency in a wild population of squirrel monkeys. The study was conducted in Eastern Amazonia, 150 km east of Belém, Brazil. I followed threegroups of monkeys in their natural habitat over six weeks (N=104 contact hours). I collected behavioral data on juveniles, adult males and adult females using focal animal sampling, recording the individual’s activity and the age-sex class of its nearest neighbor within a 5-meter radius. I also recorded all observed instances of UW, noting the individual’s nearest neighbor, temperature and humidity, social context, and potentially stressful events prior to the UW. Additionally, I recorded the temperature and humidity every 30 minutes while in contact with the study groups. Average daily rate of UW was 1.56 instances/hour (N=165 observations). Juveniles engaged in UW most frequently, accounting for 59% of observations, followed by adult females (23%), and adult males (18%). No correlation between UW frequency and temperature or humidity were found. These preliminary results indicate that UW is mostly performed by juveniles, and therefore not likely to serve a function in sexual communication. Furthermore, the hypothesis that UW acts as a self-cooling mechanism in high temperatures and low humidity was not supported. Further analyses will examine additional factors that may correlate with UW, such as social stress. The results of this study will expand our insight into why UW evolved in New World primates.

A Social Network Analysis of the #Resistance Twitter Ecology

Program: Culver Fellowship
Faculty: Dr. Jose Marichal, Political Science
Student: Ryan Mundy

As a result of the 2016 presidential election, a constellation of progressively-oriented groups have emerged (Flippable, Our Revolution, Swing Left, and Indivisible) with the stated purpose of challenging or retaking institutional power.  Using NodeXL, an open-source computer program designed to explore social media network graphs, we conducted a social network analysis of these groups’ follower/following network over a three-month period (March to June 2017) to explore the composition and changes in the network dynamics of each organization.  We look at the network compactness, cluster formation, ratio of strong/weak ties and emergence of structural holes for each organization. Using these measures allows us to examine differences within and between groups in online network composition and allows us to identify the level of integration of the emerging #Resistance activist twitter ecosystem. We focused in on three specific, progressive movements: #SwingLeft, #OurRevolution, and #Resistance and presented our work at the annual Political Networks Conference at Ohio State University. Using NodeXL, we found that #OurRevolution is focused on the “broken system” and mobilizing around it, while #SwingLeft is focused on electing progressives to office. Finally, #Resist serves to simply express general anti-Trump outrage. Furthermore, the network topology, which is the arrangement and interactions of users was unique for each movement. Our research is still a work in progress that we plan to continue into the fall. For the future, we hope to scrape our own social media network data using a program called Anaconda, rather than relying solely on the often limited NodeXL. In addition, we plan to contact the progressive organization Indivisible, to explore how they interact with other progressive communities and their best strategies for mobilizing voters using social media data.

Characterizing Ligand Binding of MAL-TIR Domain Protein within the Innate Immune System TLR-4 Pathway

Program: CLU STEM Research Abroad
Faculty: Dr. Mehdi Mobli, Biology
Student: Julie Martinez

Inflammation is a necessary element in the mammalian system’s immune response, however, in aberrant instances, it can be counterproductive and potentially harmful if excessive or uncontrolled. Thus, this research aims to develop anti-inflammatory drugs with the potential to inhibit MAL protein in the TLR4 inflammatory pathway, to ultimately hinder a counterproductive inflammatory response from occurring. NMR spectroscopy data was derived for resonance assignment utilized in tracing molecular shifts of MAL in the presence of ligand and MAL in the absence of ligand, seeking to characterize ligands that bind to note inhibitory effects on MAL filament formation. Potential chemical shifts were determined from an overlay of spectra featuring MAL in the presence of ligand and MAL in the absence of ligand, coordinating amino acid residue shifts to the corresponding residues on MAL’s protein structure. Clusters of chemical shift on MAL protein structure indicated a potential ligand-binding site, after which inhibitory effects must be observed to further develop the site as an inhibitory site for the development of anti-inflammatory drugs. Currently, ligand-binding characterization is underway to observe such inhibitory effects on MAL protein, with the expectation of developing drugs that may hinder the TLR4 inflammatory pathway ultimately restricting counterproductive inflammation.

Symmetry in the Sit-to-Stand Movement and the Role of Hip Muscle Strength in Middle Aged Adults

Program: Biomechanics
Faculty: Dr. Michele LeBlanc, Exercise Science
Student: Anthony Enzo Seddio

Transitioning from a sitting to a standing position is one of the primary movements of daily living which is performed multiple times each day. This movement becomes more challenging over time due to the strength declines associated with the normal aging process.  Previous studies have shown that the sit-to-stand (STS) movement is mechanically demanding, requiring more lower extremity joint torque and range of motion than walking or stair climbing (Berger et al., 1988). A lack of symmetry in movement has been linked to increased injury risk.  The purpose of this study was to investigate the symmetry of movement in the STS in middle aged adults and to determine if hip muscle strength plays a role in any identified asymmetry.  Healthy adults between the ages of 47 and 70  (6 males, 16 females) performed several STS movements while eight Vantage 5 cameras (200 Hz) and two Kistler force plates (1000 Hz) collected body landmark and 3-dimensional ground reaction force data, respectively. Subjects were asked to stand in front a chair of standard height (46 cm) while each foot was on a separate force plate. They were then instructed to take a seat on the chair and to sit comfortably.   Once they had been still for several seconds, they were asked to stand up and remain standing for 5 seconds. Once at least three successful trials were completed, hip muscle strength in the sagittal and frontal planes was measured using a handheld dynamometer (Lafayette Instruments).  Peak hip adduction, abduction, flexion and extension forces were identified for each subject.  Peak and average ground reaction forces (vertical, anteroposterior, mediolateral), joint ROM, and peak and average joint torques were computed using Nexus 2.5 software. Comparisons were made between the subjects’ preferred and non-preferred legs, as well as between the stronger and weaker leg, as determined by the strength measures. Leg preference was determined by each subject’s lead leg when initiating gait. Statistical significance was determined using dependent t-tests with SPSS version 25 (p ≤ 0.05).

Examining the Causal Relationship between Parasocial Relationships and Body Image

Program: Undergraduate Research
Faculty: Dr. Ariana Young, Psychology
Student: Kirby Sigler

Much research demonstrates that media ideals (i.e., thin female media figures and muscular male media figures) have unfavorable effects on men and women’s body image. However, the impact of having a parasocial relationship (PSR; a one-sided psychological connection with a media figure) on body image is much less clear. Whereas some correlational research shows PSRs are associated with worse body image, other experimental research demonstrates that PSRs contribute to better body image. The purpose of the present study is to identify the causal relationship between PSRs and body image. We argue that individuals who experience body dissatisfaction may be especially drawn to celebrities with whom they have a PSR (because it makes them feel better about their own bodies). To examine this idea, we conducted an online study with a community sample (n = 163) in which we first manipulated participants’ body image. Specifically, participants were randomly assigned to write about a part of their body they disliked vs. a control topic. Following, participants indicated their interest in various pass-time activities, including celebrity-related activities. Results revealed that participants who were led to experience body dissatisfaction reported significantly more interest in PSR activities (e.g., “read the latest news and updates about your favorite same-sex celebrity”) compared to those in a control condition, p = .05. However, this increased interest did not extend to non-favorite celebrity activities (p = .39) or other control activities (e.g., playing computer games; p = .12). Overall, these results suggest that the association between poor body image and PSRs is due to people with poor body image being drawn to their favorite celebrities. Indeed, people may seek out their PSRs when they feel bad about their bodies in an effort to feel better about themselves. This research is important because it helps provide a better understanding of the impact of celebrities in people’s everyday lives.

10:30am - 12:00pm
Poster Session

Location: Ahmanson Patio

Interactive Poster Session

Mechanistic Model for Raindrops with Persistent Superterminal Velocity

Program: Swenson Science Summer Research Fellowships
Faculty: Dr. Christopher Brown, Mathematics
Student: Jennifer Anderson

As a raindrop falls through the air, it is pulled down by gravity while it is being pushed up by air resistance. When these two forces are equal, the drop will travel at the same velocity for the rest of its fall. This velocity is called the terminal velocity. One model for terminal velocity shows terminal velocity as a function of volume when the drop is treated as a perfect sphere. It has recently been discovered that certain small raindrops can travel faster than their estimated terminal velocities. These raindrops are described as superterminal. An object is superterminal when its speed is at least 30% greater than its estimated terminal velocity. One of the hypotheses for why this happens is that large drops change shape and burst apart, forming small drops. These small drops move at the same velocity as the original large drop, causing them to be super terminal. The purpose of this research was to create a numerical model for raindrops falling to test the different hypothesis for why superterminal raindrops exist. To create this model, we used the programming language Octave. Our program is made up of functions that allow us to manipulate the variables. These functions allow us to calculate the change in the physical properties of the raindrop. The program allows the drop to fall, gain speed, reach terminal velocity, change shape, and break apart. Our results show how the system changes as you allow more variable to interact with one another. When we allowed the drop to change shape, we were able to see how the velocity and the terminal velocity changes concerning time. This research is important because the way we measure rainfall is from a raindrops velocity. Drops with large velocities are thought to be large drops. This means when measuring the large velocity of a small superterminal raindrop, we are estimating it as a large drop. This causes us to overestimate the amount of rainfall we are receiving.

Using Aziridines in Ring Expansion Reactions

Program: John Stauffer Research Fellowships in the Chemical Sciences
Faculty: Dr. Jason Kingsbury, Chemistry
Student: Anthony J. Albert

            Aziridines are 3-membered, nitrogen-based heterocycles with a high degree of ring strain. This makes them susceptible to ring expansion reactions, such as aza-Claisen rearrangements. Aziridines are useful starting materials when synthesizing larger six and seven membered nitrogen-based heterocyclic compounds. We tested three different methods to prepare complex aziridines in an attempt to design a general procedure to convert vinylaziridines into larger nitrogen-containing heterocycles, via a 3-aza-Claisen rearrangement.

            In the first method, we epoxidated, esterified and azidated cinnamic acid. While successful in producing aziridines, this approach resulted in low overall yields. In the second method, an aziridine dimer was reacted with different Wittig salts and gave three different desired aziridines in moderate yields. However, the overall scope and generality of this method are limited, as only two aziridine dimers are readily available. The most general procedure we explored was a Mitsunobu cyclization of N-trityl 2-phenylglycinol, which furnishes 2-phenylaziridine in two steps. However poor conversion and a difficult separation of the unreacted starting material from the tritylated aziridine complicate this method.

            While all of the methods outlined above suffer from at least one major drawback, results point to the Mitsunobu cyclization as the most efficient method for preparing wide arrays of aziridines. If the desired aziridine can be manufactured from an available aziridine dimer, that would be preferred as it does not involve a difficult separation or a protecting group. Next, I hope to acylate 2-phenylaziridine using a well-known procedure already successful in the case of three other aziridines. By then subjecting the acylated aziridines to a Takai olefination, the desired 3-aza-Claisen rearrangement should be promoted, allowing for the study of the regioselectivity and stereoselectivity of the rearrangement, with direct application toward medium ring amine systems.

Does California’s Agriculture Respond to Changes in Reservoir Levels and Reservoir Capacity?

Program: Overton Summer Research Program in Economics
Faculty: Dr. Matthew Fienup, Economics
Student: Alanna Bauman

For much of the State of California, 2017 marked the end one of the worst droughts in the State’s history, during which water stored in the State’s two largest reservoirs declined to less than 30 percent of capacity. Agriculture in California is a 43-billion-dollar industry and generates 100 billion dollars of total economic activity. Researchers at UC Davis estimate that the total cost of the drought to all economic sectors in 2015 to be 2.74 billion dollars and a decrease of total job losses to be 21,000 (Howitt et al., 2015). Water usage in California is split between three sectors, 50 percent is for environmental usage, 10 percent is urban, and 40 percent is for agriculture. Additionally, 40 percent of the total land usage in California is designated to agriculture uses; crop and livestock. The focus of this research is to examine the response of California’s agricultural output to changes in reservoir levels and total reservoir capacity. Using data from 1954-2007, I use regression analysis to estimate the determinants of the total crop production with controls for total reservoir capacity state-wide and individual reservoir levels. I find evidence that total reservoir capacity and reservoir levels have a significant impact on total agricultural output. Climate models predict increased variability in weather patterns in the decades ahead, meaning more frequent periods of drought punctuated by large deluges of precipitation. The results of this empirical analysis suggest that in order to maintain current agricultural output levels, without additional reservoir capacity, California needs to improve infrastructure in current reservoirs, operate dams more efficiently, and better deal with storm runoff. 

The Influence of Climate Change-Induced Temperature Increases on the Symbiotic Relationships of Anthopleura elegantissima and Aiptasia pallida

Program: Swenson Science Summer Research Fellowships
Faculty: Dr. Andrea Huvard, Biology
Student: Samantha Bever

In 2007, the IPCC predicted temperature increases for the year 2100, with 1.1-2.9°C and 2.4-6.4°C being the lowest and highest ranges hypothesized. Climate change models such as this are important tools to predict the effects of these increases on marine organisms, such as the anemones: Anthopleura elegantissima (temperate) and Aiptasia pallida (tropical), who share symbiotic relationships with intracellular algae (zooxanthellae). The goal of this study was to test how the predicted temperature range extremes would influence the expulsion of algae from their hosts, and if these results would differ between temperate and tropical species. It was hypothesized that each °C increase would cause bleaching in both species and that Aiptasia would be more affected due to their less stable symbiont populations. Five randomly selected subjects from each species were taken from control populations of n=20 anemones and used in temperature threshold tests. This process was repeated with another three anemones to determine algae control counts. Three Aiptasia and Anthopleura were then subjected to 21 days of continuous temperature increases, using a temperature range of 1-7°C that combined and simplified the original two ranges. The species sustained 1°C increases every three days with zooxanthellae counts performed every other day. It was found that Anthopleura’s algae continually declined throughout the experiment. Algal cell counts compared on days 1 and 21 had a p-value of 0.050066. Comparisons of days 1 and 6 (the lower range) were found insignificant. Aiptasia had no correlation between temperature increase and algal cell decrease for either range. In conclusion, Anthopleura showed continuous bleaching and one p-value near .05 while Aiptasia did not, despite their less stable symbiont populations. This suggests both ranges influenced expulsion in the temperate species, but the lower range did not influence both species to the extent predicted; therefore, Anthopleura’s algal expulsion may be more influenced than Aiptaisa’s if temperatures increased to predicted values. 

RNA Sequencing and Gene Expression Profiling of HCV Infected Monocytes Grown in Adult Bovine Serum

Program: Swenson Science Summer Research Fellowships
Faculty: Dr. Dennis Revie, Biochemistry
Student: Nicholas Bui

RNA from Hepatitis C Virus infected and uninfected monocytes were grown in fetal bovine serum (FBS) has previously been sequenced. This study intends to determine which genes are differentially expressed from monocytes grown in adult bovine serum (ABS). To do this, mRNA from infected and uninfected monocytes grown in ABS was purified and prepared for sequencing using the Illumina TruSeq Stranded mRNA Library Preparation Kit. Samples were sequenced using the Illumina Miseq sequencer. DNAStar was used to align and analyze the RNA reads, estimate RPKM of reference genes, and assemble novel transcripts.  Each sample yielded over 26.4 million high-quality reads. The genes that changed most significantly were determined: Two genes that encode for proteins related to the Innate Immune System Pathway were upregulated, SPPL2B (14989.8-fold) and BTN2A1 (9684.6-fold). SPPL2B encodes Signal Peptide Peptidase Like 2B and BTN2A1 encodes Butyrophilin Subfamily 2 Member A1. Two genes that are involved in structural cell changes and migration were also upregulated, PLXNA1 (2030.7-fold) and CROCC (1983.2-fold). PLXNA1 encodes Plexin A1, and CROCC encodes Ciliary Rootlet Coiled-Coil Protein. CDKN2A was upregulated (19343.3-fold) and encodes Cyclin Dependent Kinase Inhibitor 2A which assists in the regulation of cell proliferation. ACYP1 was upregulated (17801.0-fold) and encodes Acylphosphatase 1, an enzyme related to pyruvate and naphthalene metabolism pathways. IRS2 was upregulated (2895.1-fold) and encodes Insulin Receptor Substrate 2, a cytoplasmic signaling protein involved in insulin mediated cell processes. MNT was upregulated (6599.0-fold) and encodes MAX Binding Protein that binds DNA as a heterodimer with MAX to repress transcription.   HGH1 was upregulated (9837.7-fold) and encodes HGH1 homolog (p < 0.01, t > 671.07). These same genes were also most statistically significant for the samples from monocytes grown in FBS. These results suggest that the serum does not affect gene expression for HCV infected monocytes.



Dual Platform Metabolic Characterization of Excreted Oxidative Phenolic Compounds as Potential Biomarkers

Program: John Stauffer Research Fellowships in the Chemical Sciences
Faculty: Dr. Grady Hanrahan, Chemistry
Student: Jeffrey Burr

Previously, we demonstrated an untested link between phenolic compounds and the formation/electrophoretic separation of stable urinary metabolites. However, there is a paucity of information as it relates to full chemical characterization and dose- and time-dependent urinary metabolic screening. By exploring the relationship between structural characteristics and toxicity response, this study will lead to a greater understanding of underlying disease mechanisms. Methodologically, we will use a combination of gas chromatography/mass spectrometry (GC-MS) and capillary electrophoresis (CE) to separate and structurally characterize a variety of phenolic-based compounds for use as potential biomarkers as a result of short- and long-term industrial exposure. Integral steps to this investigation will include instrument optimization, sample preparation, and subsequent data analysis. As eluded to previously, we expect this new study will guide the development of a comprehensive approach towards dose- and time-dependent urinary metabolite screening. Examination for potential phenolic biomarkers will aid in understanding underlying disease mechanisms and potential ways to avoid/limit exposure. Potentially exposed and non-exposed individuals will eventually be recruited to participate in this follow-up study once Institutional Review Board (IAB) approval is obtained. Attempts to achieve optimal separation of different phenolic compounds by capillary electrophoresis were carried out through the implication of multiple surfactants combined with either a phosphate or borate buffer. When running standard solutions containing multiple phenolic compounds, we have yet to see an optimal separation of each individual compound. However, an optimal separation of pentachlorophenol standard was achieved through a borate/Sodium Dodecyl Sulfate (SDS) buffer solution. Further analysis by altering the pH and buffer/surfactant concentrations will be performed until optimal separation of each compound is achieved.

Investigating the Presence and Quantity of Immune Cells in Canine Tumors

Program: Swenson Science Summer Research Fellowships
Faculty: Dr. Chad Barber, Biology
Student: David "Andrew" Crago

The microenvironment of canine tumors requires further study, including knowing the immune cells that are present. Therefore, the project aims to investigate the quantity and presence of immune cells in canine tumors. It is hypothesized that the quantity of immune cells present in canine tumors is similar to comparable human tumors in the literature.  Seven fresh canine tumors were received from veterinarians. A portion of all tumors were frozen for later DNA analysis. Six tumors were then processed to achieve a single cell suspension. A portion of the processed tumor cells were grown in cell culture to attempt to produce an immortal cell line. It was found that cells from a canine papilloma tumor grew in culture till passage four with loss of contact inhibition but died due to possibly reaching the Hayflick limit. All other primary cells from the tumors died in culture and did not become immortal cell lines. A different portion of the processed cells in four of the tumors were stained with fluorescently labeled antibodies (anti-CD45R, -CD8a, or -F4-80) to identify immune cells.  The stained cells were analyzed by flow cytometry to determine cell types present. It was found that blood-related immune cells were identified in four of the labeled tumor samples. The prominent cells identified were hematopoietic in one tumor (2.47% CD45R+), Macrophages in one tumor (3.94% F4-80+), and Cytotoxic T-cells in two tumors (6.70% CD45R+CD8a+, 4.94% CD8a+). It is concluded that Cytotoxic T-cells are an important part of the canine tumor microenvironment. The project will continue to test more canine tumors to gather more data, and then the data will be compared with published data on comparable human tumors from the literature. By understanding the immune cells in tumors, both veterinarians and physicians can better understand the role of the immune system in cancer.

The Biochemistry and Bioinformatics of Selected Ocular Proteins from Bos Taurus

Program: John Stauffer Research Fellowships in the Chemical Sciences
Faculty: Dr. Robert Richards, Chemistry
Student: Kacey Egusa

     Infectious bovine keratoconjunctivitis commonly referred to, as “pinkeye” in cattle, is an ocular disease that is endemic in cattle herds worldwide. Our previous study determined the bacteria present in each condition whereas this study focused on identifying genes known to cause infection in the healthy versus infected cattle and to learn more about the virulence mechanisms used by pathogenic bacteria.               

     This study involves the verification and use of MG-RAST, a web application that analyses metagenomes. Two methods were used to establish the reliability of the data produced by MG-RAST. The first technique exploited a sample that contained only E. coli as the sole bacteria identified. By utilizing the KEGG database, it was possible to identify more than 61% of the putative proteins included in a number of known pathways.  The second technique utilized the glycolysis/gluconeogenesis pathway in KEGG.  Comparing all the samples to the pathway more than 85% of the putative proteins were identified (n=20, SD = 1.8).  These two techniques suggested that MG-RAST identified a majority of the proteins present.

     The second objective was to identify functional differences in healthy versus infected samples to investigate how bacteria infect calves. Commonly known virulent factors and pathways were considered.  Our investigation showed a significant lack of genes from the type III secretion system and the two-component system genes AgrA, SaeS and SaeR in healthy samples even though some contained known pathogens.  These missing virulent factors, required for pathogenicity, suggest a possible explanation of the noted avirulence. This study of virulent factors derived from functional metadata can be used to gain a better understanding of the biological interactions and virulent mechanisms in a mixed environment of organisms.

Fighting Fake News One Article at a Time

Program: Swenson Science Summer Research Fellowships
Faculty: Dr. Chang-Shyh Peng, Computer Science
Student: Cole Epel

Introduction: With the rise of independent news sources, the concept of “fake news” has always been an active threat. The goal of this project is to attempt to understand how fake news works since, as far as its appearance online is concerned, it’s a relatively new issue. The hypothesis is that if multiple news stories on a central topic are accurate, they should share similar information. The program designed over the course of this Summer was meant to test for these very similarities by using’s model for comparing papers against its database.


Current Project Status: Currently, it can extract any given news story from the HTML code embedded in a web page using an API called “Jsoup.” With this story, the program can provide a number which represents the amount of similarities between one story and another with an API called “java-string-similarity.” The specific method for comparing stories is known as a "Normalized Levenshtein Comparison." This gives us the percentage of operations required to turn one String into another String, producing a percentage of “differences” between the two. By subtracting one from that percentage, the number provided represents the amount of similarities between the two stories. By running this comparison multiple times with one story against a series of stories in the “database,” the average of those percentages was used to determine whether or not a story could be “fake.”


Conclusion: The threshold used for these tests was if the percentage was within the range of 50% or above, the story was most likely real or related to the articles in the database. These results, with the proper conformation, could provide us with a new understanding of how fake news is identified.

Vashon Island Suite

Program: Summer Undergraduate Research Fellowships (SURFS)
Faculty: Dr. Mark Spraggins, Music
Student: Grant Escandon

Vashon Island is a large island in the Puget Sound with views of Mt. Rainier and many scenic roads. Residents enjoy a rural atmosphere, with several small towns and schools. This new piece for wind ensemble offers an impressionistic view of island life, with four movements, Sealth, Milky Way, Lisabeula, and Fall, serving as distinct sound-paintings of different aspects of the island. The composition process can be broken into four stages, consisting of sketching, writing short scores, orchestrating and engraving. Sketching was the initial stage of writing. It involved notating any harmonies, textures, structural ideas, or melodic fragments that came to mind. Once sketches for each movement were complete, they were refined into short scores, which are versions of the final music reduced to around four staves. The staves were separated by layers of harmony, melody, and textural devices so that they could easily be isolated and refined during orchestration. Through the orchestration stage, each voice was assigned to an instrument. Careful consideration of instrument ranges, timbre, transpositions, and overall color and balance of the ensemble took place in this step. The engraving was done in MakeMusic’s Finale software. Engraving results in the production of a score, which the director will conduct from, and parts, which each player will receive. Although there were distinct stages to composing, it is important to note that it was not always a linear process. “Milky Way” came very easily and quickly and was the first movement to enter orchestration, far ahead of the others. “Lisabeula” was more difficult to write, and took quite some time to complete; with many drafts before it’s final form came into shape. There were also movements with complete sketches that were set aside simply because they did not fit the mood of the piece as a whole.

This piece will be submitted to regional conferences through SCI, Society of Composers, Inc., with the hopes of being performed throughout the US. 

Metrization Theory in Topology

Program: Swenson Science Summer Research Fellowships
Faculty: Dr. Nathan Carlson, Mathematics
Student: Ryan Fisher

In mathematics, a topological space is an abstract setting defined by open sets.  Over the past century or more an ongoing project in topology has been to determine what types of topological spaces are in fact metric spaces. These are spaces with a metric, that is, a function that measures the distance between two points. A space is metrizable if it has a metric that generates its topology. Indeed, historically the notion of a topological space grew out of a need to generalize metric spaces into a broader setting where there might not be a metric that agrees with basic properties a distance function should have. One set of properties that require a space to be metrizable is second countable and T­3. This was originally proved in 1925 by Urysohn and is known as the Uryzohn Metrization Theorem. My goal for my summer research was to prove this theorem. The process of proving this theorem started with background readings in general topology followed by focusing specifically in metrization theory. These readings were supplemented with a discussion with my mentor, Dr. Carlson, and another math faculty, concluding in the formal proof of the Uryzohn metrization theorem.  This process took about eight weeks of rigorous study.

The Uryzohn metrization theorem specifically states that a T 3-second countable topological space is metrizable. This set of properties is only one combination that leads to a topological space being metrizable, and the larger study of metrization theory is about what properties require a metrizablity. Though I proved the Uryzohn metrization theorem specifically I encountered many others with a large variety of properties leading to metrization, showing how the field of metrization theory has been studied for over a century and continues to be studied today. This introduction to topology as a whole and metrization theory more specifically showed me the inner workings of metric theory, which I studied previously in one of my courses.

Effect of Hip Strength Asymmetry on Functional Lower Body Gait Asymmetry in Healthy Middle-Aged Adults

Program: Swenson Science Summer Research Fellowships
Faculty: Dr. Michele LeBlanc, Exercise Science
Student: Spencer Gottmer

The use of one side of the body more than the other can lead to deterioration of the overused side. Many studies have shown that asymmetrical movement and muscle use can cause injury. A previous study focusing on injury prevention after prior recovery from anterior cruciate ligament strains or tears found that one of the main four predictors of a secondary injury is lower body asymmetry; specifically, sagittal-plane knee joint asymmetries (Hewett et al., 2013). Walking is one of the most commonly used movements in day-to-day life. Therefore, if asymmetry is present in gait, there is a high risk of overuse injury. The purpose of this study was to determine whether hip strength asymmetry, in injury-free middle-aged adults, affects gait asymmetry. Twenty-five middle aged adults walked 10 meters at a self-selected pace several times with sixteen reflective markers affixed to key body landmarks. Eight Vantage 5 cameras (200 Hz) and two Kistler force plates (1000 Hz) collected three-dimensional coordinates and ground reaction forces, respectively.  Several walking trials were collected to ensure that foot contact was fully made with the force plates without any adjustments to the subjects’ natural gait patterns. Hip joint muscle strength was measured using a handlheld isokinetic dynamometer (Lafayette Instruments).  In particular, peak joint flexion, extension, adduction, and abduction were used to determine hip strength in the sagittal and frontal planes.  Hip, knee, and ankle joint angles, ROM, and peak and average joint torques were computed using the Nexus 2.5 software.  Ground reaction forces (vertical, anteroposterior, mediolateral), and lower body kinematics and kinetics were compared between the preferred and non-preferred legs and between the stronger and weaker legs using a dependent t-test. Leg preference was defined by the lead leg when walking naturally.  Statistical significance was determined with SPSS version 25 (p ≤ 0.05).



Enantiospecific Azepine Ring Formation via Aziridine Rearrangements

Program: John Stauffer Research Fellowships in the Chemical Sciences
Faculty: Dr. Jason Kingsbury, Biochemistry
Student: Karam Malki Hajjar

Azepines are unsaturated heterocycles of six carbons and one nitrogen atoms. Various methods have been developed in total syntheses to construct azepine rings; however, limited focus has been given to the stereochemistry of the medium ring azacycles. Working with Dr. Kingsbury, our efforts have been towards developing phenyl substituted and optically active N-acyl vinyl aziridine as a substrate for forming new stereocontrolled azepine ring via [3,3]aza-Claisen rearrangement. Aza-Claisen rearrangements are [3,3]-sigmatropic rearrangements on N-allyl vinyl-amines that have shown to require temperatures as high as 250o C. Incorporation of a ring strain element to facilitate the rearrangements represents an interesting strategy that has not been investigated thoroughly. The Synthesis of a phenyl-substituted azepine can be accomplished when 2-phenyl-3-vinyl aziridine undergoes installation of a carbonyl group, either formylation or acylation, followed by Tandom-Takai Olefination to reduce the carbonyl functional group to an alkene. Following two formylation methods on secondary amines, it was shown by NMR analysis that the aziridine ring was not conserved and had undergone undesired rearrangements, while the methods for acylation formed a secondary acylated amine with a conserved aziridine ring. Formylating the secondary amine is a challenging step in the mechanism because of ring stain that the azirdine upholds and the electron density that could be donated from the vinyl group via resonance. Studying aziridine rearrangements to azepines could help establish an enantioselective total synthesis for many useful drugs like Proheptazine, an analgesic opioid that belongs to the 4-phenylazepine family. 

Characterization and Analysis of Integrins on Immune Cells found in Canine Tumors

Program: Summer Undergraduate Research Fellowships (SURFS)
Faculty: Dr. Chad Barber, Biology
Student: Mary Jarjour

The purpose of this study is to identify and analyze integrins found on cytotoxic T cells, macrophages, and tumor cells present in canine tumors. Additionally, we aim to identify roles of these integrins in immune cells as well as other cells found in the tumor. Based on previous studies, I predict that the canine integrin aLb2 will be found on cytotoxic T cells and the integrin aMb2 will be found on macrophages in canine primary tumors analyzed. Primary methods for this study include cell culture and flow cytometry. Tumors received from veterinarians are rinsed with basal media and disaggregated; half the cells are seeded for cell culture. The remaining half is used for flow cytometry. We stained the cells with fluorescently-labeled antibodies that bind to specific integrins. The cells are run through the flow cytometer to detect integrin expression of the cells. We have been using antibodies both specific to canine species, but also some that are cross reactive for the canine. We have seen the expression of the aL subunit with two tumor samples. In previous studies, we have found expression of the a6 subunits on the canine cancer line, D17. We aim to perform inhibition assays with integrin blocking antibodies or peptides in vitro to assess the role in survival, migration or adhesion in the tumor microenvironment. By learning more about the integrins expressed on cytotoxic T cells and macrophages in canine tumors, we can make predictions about the role that the integrins play in cancer progression.

The effects of glutamate and the glutamate antagonist, CNQX, on the feeding behavior of Lymnaea stagnalis and Helisoma trivolvis

Program: Swenson Science Summer Research Fellowships
Faculty: Dr. Kenneth Long, Biology
Student: Kailey Jordan

Studies of the molluscan nervous system have provided fundamental knowledge of neural interactions. Glutamate is a common neurotransmitter in both vertebrates and invertebrates. I used both isomeric forms of glutamate, at specific concentrations, to assay its effects on feeding behavior of the freshwater snails Lymnaea stagnalis and Helisoma trivolvis. The glutamate receptor antagonists, CNQX and DNQX, were studied to test their ability to block the effects of glutamate. 

A “two-choice” test was used to determine the threshold concentration of L-and D-glutamate required to inhibit feeding behavior. Snails were fasted for 24 hours and tested individually using Sylgard-lined containers, which included two equidistant 20 mM maltose pellets. Twenty snails were tested at each concentration of L- and D-glutamate. Concentrations of L- and D-glutamate tested ranged from 10-6 M to 10-2 M. The number of snails feeding on pellets was recorded every 5 minutes over a 1-hour interval. The lowest concentration of L- or D-glutamate that produced significant (p<0.05) inhibition of feeding was 10-4 in both species.

To assess L- and D-glutamate receptor affinity for antagonists, a pre-treatment of CNQX or DNQX was prepared. Twenty L. stagnalis or H. trivolvis were placed in 10-4 M CNQX or DNQX for 15 minutes and then tested as described above. This procedure did not inhibit feeding. Snails were then tested using equal concentrations (10-4 M) of CNQX and L-or D-glutamate. Means for control and experimental assays were calculated. A 1:1 ratio of CNQX to L-glutamate blocked the inhibition of feeding in L. stagnalis; results were not significant using H. trivolvis.

Results of my study demonstrated that both L- and D-glutamate inhibited feeding in L. stagnalis and H. trivolvis in a dose-dependent manner and that the glutamate receptor antagonist CNQX blocked the inhibition of feeding in L. stagnalis. Further research will be conducted to understand how glutamate inhibits the feeding system of freshwater snails using electrophysiological techniques.

Strategies for Multi-Core Joins

Program: Swenson Science Summer Research Fellowships
Faculty: Dr. Graham Matthews, Computer Science
Student: Kevin Hwa Lee

An essential operation when querying data is joining, where data from one source is combined with data from another. Joining data from multiple sources allows us to answer questions that cannot be answered from a single source alone. Making joins execute efficiently is a central problem in database systems. Existing strategies for joins are optimized for single-core CPUs, specifically to minimize time to first result or memory usage. However, modern hardware features multi-core CPUs – our research explored effective join strategies in this context.

A database query can be represented as a binary tree, where each node is a data processing operator. Two operators of particular interest are receives, found at the leaves of the tree and used for retrieving data from data sources, and joins, found as internal nodes and used for combining data. Data starts at the receives and flows through the tree to the root. Trees can be executed in more than one way. We first developed an execution strategy, the Queuing Strategy, which employs a standard “threads-and-queues” approach. Each operator in the tree executes as a thread and between the operators are buffers allowing one operator to pass data to another. The queueing strategy functioned as a baseline against which we can measure other strategies. We then developed a second strategy, the Phased Strategy, where trees are executed in two phases. First, all join and receive operators on the left of the tree are executed in individual threads. Second, a group of worker threads is used to take data from the rightmost receive in the tree and push it through the tree to the root.

            The two strategies were benchmarked. We expected that the phased strategy would outperform the queueing strategy. Our results, however, show that: a) the phased strategy only outperforms the queueing strategy on Linux – on OS X the two strategies perform comparably: b) the queueing strategy exhibits greater performance variability than the phased strategy on all platforms.

Robotic Vision and Spacial Awareness

Program: Swenson Science Summer Research Fellowships
Faculty: Dr. Craig Reinhart, Computer Science
Student: Damian Mirizzi

In the case of a large scale earthquake, it will be very difficult to search multiple destroyed buildings for the injured. To solve this, I wanted to create a device that can search rooms; draw pictures of the room dimensions and the objects in the room without an operator. To perform this, I constructed a robot with two computers aboard. One computers job was to do simple tasks like changing the wheel speed. It also used sonar and accelerometer to sense its environment. It did this by coupling the accelerometer’s compass and the sonar distances to triangulate the location of an object relative to the robot. The second on board computer took photos and read the sensor data from the first computer; it did this so that it could draw its surroundings and distances creating a blue print of the room and the objects in it. Each robot will be connected to the main server, the server is in charge of all the picture analytics, so when the second computer takes a picture, it sends it to a server.  The server will then reply with what object was in the photo. The server is using a Tensorflow backend to do the image classification. It is a multi dimensional convolutional neural net. The purpose of the distributed design is so that the robots can stay cheap with only a few small computers on board. Moreover, leave all the tough computation to the server that can serve hundreds of thousands of computers at a time.  My results to date are that I can have my robot monitor the room and place objects in the room, The accuracy is +/- 10 % on the exact location and the classifier is 60 percent. I would like to make more improvements and to have the robot move around more, to better represent a room. I hope that technology like this could be in the houses of multiple people, It would be vital for saving lives in an emergency situation. It can also be duel purpose like a vacuum or be used to do other things around the house.  

Analyzing the Gut Microbial Community of Green Sweat Bees, Agapostemon spp., from California Coastal Mainland Coastal Mainland

Program: Swenson Science Summer Research Fellowships
Faculty: Dr. Theresa E. Rodgers, Biology
Student: Asia Moore

Pollination of many crops in modern agriculture relies on honey bees, which were introduced to the American continents by European immigrants. Within the last decade, worldwide honey bee populations have been threatened by Colony Collapse Disorder (CCD), in which the majority of worker honey bees abandon the hive, leaving behind the queen and immature bees to die. Although no one specific cause is attributed to CCD, many factors including pesticides, pathogens, and nutrient shortage contribute to the colony decline. Several studies have linked honey bee health to their gut microbial community (microbiota). The decline in honey bee colonies highlights the importance of native bees for crop pollination. However, research on native bees is lacking. North American bumble bees, the most studied of native bees, have gut microbiota distinct from that of honey bees. We expect other native bees, such as green sweat bees, have gut microbiota distinct from both honey and bumble bees.

Native to the American continents, green sweat bees are primarily solitary bees that pollinate a variety of plants and burrow into the ground to form nests. This study is the first to explore the gut microbiota of green sweat bees (Agapostemon spp.) native to the coastal Southern California mainland. We collected female and male green sweat bees from the Ventura Botanical Gardens, Ventura, CA. The bees were chilled on ice, anesthetized with carbon dioxide, then dissected to extract the fore-, mid-, and hindgut from which DNA was extracted. We performed polymerase chain reaction (PCR) to tag individual samples and amplify small subunit ribosomal DNA sequences of Bacteria and Archaea. We are using the Illumina MiSeq Next-Generation Sequencer to sequence all samples and QIIME software to assess sequence quality, determine phylogenetic identity, and analyze α- and β-diversity. We expect the gut microbiota to be distinct between native green sweat bee and honey bee populations.

The HGCAL-CMS Upgrade and Strategies for improving SUSY Data Analysis Searches

Program: Swenson Science Summer Research Fellowships
Faculty: Dr. Sebastian Carron Montero, Physics
Student: Johanna Paine

At the center of heavy particle measurements in the Compact Muon Solenoid (CMS) experiment at the Large Hadron Collider (LHC), and especially in searches for new and yet undetected heavy particles such as supersymmetric particles and other dark matter candidates, lies the need to identify with high efficiency secondary vertices originating from decays into b-quark jets (b-tagging). Our group at CLU collaborates with the high energy experimental group at UCSB in searches for yet undetected supersymmetric particles. In particular, supersymmetric top quarks pair production is expected to produce secondary b-quark decay vertices. Given that the expected production rates for our signal process is several orders of magnitude below background production rates, maximizing our b-tagging efficiency is of central importance to achieving the highest possible sensitivity in our all hadronic decay channel search for supersymmetric top quarks. Here we present the results of our study, outlining the expected advantages of using a multivariate self-learning technique, a deep support vector machine learning methodology, rather than traditional event cut techniques, for identifying b-quark jets.

Development of Testing Methodologies and Performance Measurements in Prototype Hexagonal Modules for the High Granularity Calorimeter Detector upgrade for the Compact Muon Solenoid experiment at the Large Hadron Collider

Program: Darling Summer Research Fellowships for Applied Scientific Computing
Faculty: Dr. Sebastian Carron Montero, Physics
Student: William Parquette

We will present the results of our summer efforts to develop testing capabilities for silicon based pixel detector prototypes, which constitutes the essential component of the High Granularity Calorimeter (HGCAL) upgrade of the Compact Muon Solenoid (CMS) experiment at the Large Hadron Collider (LHC) in Switzerland. The HGCAL upgrade is a collaborative effort involving California Lutheran University, UCSB, CERN and other institutions to increase the discrimination resolution for hadronic and electromagnetic particle showers, as well as provide an improved determination of the energy scale for hadronic quark-gluon particle showers. This experimental signature is a fundamental parameter in searches for new, yet undetected, heavy particles at CMS, including the search for possible dark matter candidates. Over the summer we have developed mobile test-station instrumentation, software for operating the test-station and data analysis software to determine important performance parameters of the detector modules, developed collaboratively by the research consortium. We will present results including measurement of analog to digital converters (ADC) noise levels, silicon base pedestal noise levels, pn junction leakage currents, high-voltage sensor response and minimum ionizing particle projected sensitivity. These results are an important step in the development of the HGCAL detector upgrade and fundamental for our group longer term search for Dark Matter at the LHC.

Geology of the Lower Miocene Monterey Formation in Point Dume, California

Program: Swenson Science Summer Research Fellowships
Faculty: Dr. William Bilodeau, Geology
Student: Alexia Rojas

Exposures of the Miocene Monterey Formation, consisting of uplifted marine sediments, can be found in the coast ranges and sea cliffs along the southern California coast. The largest accessible and continuously exposed section near Cal Lutheran is located in Malibu at Point Dume state beach. The purpose of this study was to use clues within the formation to determine the ancient sedimentary processes that deposited the rocks and the subsequent deformation that produced the exposures found today. I investigated the lowermost section of the formation along the sea cliffs between Point Dume and Paradise Cove by measuring strike and dip, collecting samples and recording both sedimentary and deformation structures. Evaluation of the area centered on a large fault, not noted on previous geologic maps, separating two areas of different structural and sedimentological composition. Evidence of the fault displacement is exposed during low tide. While the rocks on one side of the fault dip an average of forty-two degrees the rocks on the other side dip a shallow nineteen degrees. West of the fault, the deeply dipping side, consists of a shale member interbedded with diatomite and chert with small scale brittle deformation. East of the fault, the shallow dipping side is a porcelanite member interbedded with mudstone that underwent both brittle and ductile deformation (folding). This eastern member contains many low angle thrust faults and folds that could be tied to drag-folding caused by slip on the larger fault. This study is important because much of the previous research in this area on the Monterey Formation has concerned its economic importance, i.e., oil and diatomite found in the formation, rather than structural analysis. 

Preparing not to Forget: Actions People Take to Avoid Memory Failure

Program: Summer Undergraduate Research Fellowships (SURFS)
Faculty: Dr. Andrea Sell, Psychology
Student: Lorena Rosales

Metamemory refers to the knowledge and awareness of the accuracy of memory (Flavell & Wellman, 1975). Most of the current research in this area has been done on improving memory for those who have varying degrees of cognitive impairment from aging or injury. Much of it focuses on the nature of forgetting; who forgets, how they forget, when they forget; and are done in a laboratory setting (Zhou, Lu, & Dong, 2017; Fastame, 2014; Hourihan & Benjamin, 2014). However, there are not many of these types of studies outside a lab context for healthy, functioning adults. The purpose of this study was to examine what actions and behaviors people take in everyday life (as opposed to a lab setting) to plan for potential memory errors and remember important items and events. We hypothesized that certain individuals would plan and prepare more effectively than others for potential future memory errors to prevent them from occurring. To test this idea, participants answered questions through Mechanical Turk. For example, “What do you do to remember to take medications and vitamins every day?” Participants also answered questions about how often they forget, for example, “How often do you forget to take [medications and vitamins]?” Preliminary analyses show that people often use their phone to remember an upcoming task. Furthermore, the more participants reported using their phone to remember a task (compared to other methods of remembering) the lower their total forget score, r(133)=-.172, p=.046. This correlation suggests that phones can be a helpful way to prevent forgetting. Further analysis of the data will examine the effectiveness of internal v external memory aid strategies.

Body Size, Limb Length, and Sprint Speed of Sceloporus Occidentalis in Urban and Non-urban Environments

Program: Swenson Science Summer Research Fellowships
Faculty: Dr. Kristopher Karsten, Biology
Student: Brian Song

With comparatively abnormal structural density urban environments create a relatively new micro-habitat for animals to survive in. Urban environments with its large swathes of terrestrial landscape and staggering verticality present a much different habitat than a non-urban environment. These drastically different microhabitats may lead to a very different selection pressure for animals that rely on agility. The purpose of this study was to analyze the effect that urbanization has on the sprint speed, limb length, and body size of Sceloporus occidentalis. While also looking for a positive correlation with limb length and maximum sprint speed. Lizards were collected in three locations Cal Lutheran campus, Mt. Clef hills, and Malibu Creek state park. Once a lizard was collected it was taken back to the lab to be incubated for an hour at 35˚C. Once incubated the lizards were ran on a 2m track with infrared sensors every 25cm. Each lizard was raced three times with an hour break in the incubator between each sprint. Once sprint trials were completed a picture of each lizard was taken, these pictures were later analyzed with ImageJ. This software allowed the measuring of the lizards’ individual limbs. A total of 21 urban lizards were collected, the average speed found for these urban lizardson Cal Lutheran campus was 1.822 m/s, with the fastest lizard being 3.782m/s and the slowest lizard being 0.652 m/s. The average body weight of the lizards collected on campus was 11.21g. An average speed to body weight ratio of 0.2286:1 for urban lizards was found. The data collected shows a positive correlation between body size and sprint speed. Further data collection is required with two nonurban data sets needing to be collected. If the hypothesis is failed to be nullified, it shows that urbanization causes concrete morphological changes in species that can significantly change their survival rate.

Developing and Testing a Measure of Vocal Bias in News Media

Program: Summer Undergraduate Research Fellowships (SURFS)
Faculty: Dr. Monica Gracyalny and Ryan Medders, Communication
Student: Amanda Souza

This research investigates whether people perceive bias in a voiced news report. Much of the research on detecting bias in the news media has focused on the verbal content of news stories, such as the amount of favorable or unfavorable quotes from sources or the specific language used by reporters. Few studies, however, have examined how reporters’ nonverbal cues (e.g., vocal characteristics, facial expressions, body movement) affect audience perceptions of media bias. Because nonverbal behavior is generally considered to be a more reliable and valid indication of a speaker’s true feelings, this pilot study investigates vocal tone as a bias cue in news reporting. The current study develops and tests a range of vocal stimuli representing favor or prejudice toward a particular topic through variations in speakers’ pitch, the rate of speech, and volume. Five audio recordings were created to manipulate the various vocal characteristics in a hypothetical voiced weather report, and variations were confirmed using Praat audio analysis software. Participants (n = 409) listened to five randomly presented audio recordings via Qualtrics and reported their perceptions of the strength and direction of the actor’s bias toward the weather in each recording. Preliminary results showed that the actor’s combination of high pitch, fast rate, and high volume was perceived as a little biased (M = 2.09, SD = 1.21), and between somewhat and very positive (M = 4.31, SD = .96) about the weather. The combination of low pitch, slow rate, and low volume by the speaker was rated as a little biased (M = 1.65, SD = .89), and between neutral and somewhat negative (M = 2.73, SD = .87) about the weather. Findings suggest that participants can easily identify speakers’ positive bias in manipulated audio recordings. However, a negative bias is more difficult to identify through pitch, rate, and volume alone. Future research should investigate the effects of additional vocal qualities such as inflection and pitch range in the perception of negative bias.

Adaptive Health Behaviors: The Relationship between the Behavioral Immune System and Memory

Program: Culver Fellowship
Faculty: Dr. Andrea Sell, Psychology
Student: Gabriella Steffon

The present study focuses on the behavioral immune system, which includes the behaviors that one takes in order to prevent themselves from getting sick. Specifically, this study investigates how these behaviors may affect one’s memory of environmental details. Recent evidence has suggested that certain psychological processes such as memory and categorization are used by the behavioral immune system to detect possible health threats (Miller & Maner, 2012). For example, when participants are concerned about pathogens, they categorize potential disease items faster than neutral items (Miller & Maner, 2011). Therefore, we hypothesized that people would better recall specific items that a sick individual comes into direct contact with and the symptoms they display, more so than they would recall peripheral details of the environment. To test this idea, participants (n=78) were asked to read a short story about a person at work. They were randomly assigned to an experimental condition in which the scenario included a sick co-worker or a control group in which the sick co-worker in the story was replaced with a tired co-worker. Participants then completed a short filler task to ensure that details recalled were not from short-term memory. Lastly, participants were asked to recall and describe as much of the story as possible as well as answer questions about their level of disgust and worry about getting sick. Several independent t-tests and repeated measures ANOVAs were conducted to analyze these effects. Results show that participants who read the story about the sick co-worker reported more correct symptoms that the individual in the story displayed than peripheral details. However, participants in the tired condition reported more peripheral details than correct symptoms of the tired individual. These results underscore the idea that memory may be one tool we use as part of the behavioral immune system, and can help us to avoid coming into contact with disease pathogens. 

Monkeys with an Oedipus complex? A Study of Juvenile Squirrel Monkey Aggression toward Adult Males during the Mating Season

Program: Swenson Science Summer Research Fellowships
Faculty: Dr. Anita Stone, Biology
Student: Stephanie Straw

In most primates, adult males are dominant to adult females and juveniles. Yet, in some species of squirrel monkeys, adult males are subordinate.  Previous research has documented a female’s youngest offspring (young juvenile, or J1) directing aggression toward adult males during the 8-week mating season. Also, at this time adult males display weight gain known as “male fattening,” leading to a fattened appearance in the upper body. The purpose of this study was to examine the context of juvenile-male aggressions (hereby, JMA) to test two hypotheses about why they occur. Because adult females are dominant to adult males, J1 females may perform most of the JMAs, to assert their dominance from an early age. On the other hand, due to the subordinate status of less fattened males, J1s of both sexes may perform JMAs equally. Three groups of squirrel monkeys were studied in Amazonian Brazil, over six weeks (N=101 contact hours). I conducted focal animal samples on adult males and J1s of both sexes, during which I recorded the individual’s activity and nearest neighbor within a 5-meter radius. Any interaction between adult males and J1s also were recorded as they occurred. I noted the sex of the J1, the level of fattedness of the adult male, the J1’s behavior before the JMA, whether an adult female was present, and the outcome of the event. Preliminary results show that 76% of the interactions between J1s and adult males involved aggression (N=77 interactions). JMAs were mostly initiated by male J1s (62% of JMAs, N=51). In addition, fatter males were more likely to be tolerated by J1s or to ignore a J1 when receiving aggression.These early results suggest that JMAs are not driven by J1 females trying to assert their dominance over adult males in this female-dominated species. It is possible that JMAs are a way for juveniles of both sexes to maintain their status over males who are less fattened, younger and more subordinate. My study is one of the few to investigate juvenile-male interactions in neotropical primates, which is important to better understand primate social dynamics.

Near Unit Roots: A Simulation Study

Program: Overton Summer Research Program in Economics
Faculty: Dr. Dan Hamilton, Economics
Student: Xin (Catherine) Tang

Unit roots are one of the most commonly seen characteristics in time-series economic, financial, and demographic data. The strong persistence leads to inaccurate results in regressions using OLS. Researchers have developed alternative techniques, including Fully-Modified OLS (FMOLS), to improve the accuracy, and especially, inference. However, it is extremely hard for the applied researchers to distinguish between unit roots and near unit roots. This is due to the low power in unit-root hypothesis testing. Hence, it is essential to study how well the FMOLS and the OLS estimators perform under both unit roots and near unit roots. In theory, the FMOLS estimator performs better than the OLS estimator for unit roots. Because unit roots and near unit shoots share similar properties, we hypothesized that, for near unit roots, the FMOLS estimator still performs better than the OLS estimator. 

In the research, the Monte Carlo simulation method is used to compare the performances of the OLS and the FMOLS estimators in estimating the univariate long-run equilibrium relationships with unit roots and near unit roots. The results indicate that the FMOLS estimator improves significantly upon the OLS estimator for both cases overall, particularly with respect to inference. Remarkably, the FMOLS estimator performs surprisingly well with near unit roots despite being theoretically disadvantaged. We find that different sample sizes, different distributions used in the DGPs, and different types of errors included in the co-integration with unit roots and near unit roots can vary the results accordingly. The results also show that the gains from using the FMOLS estimator are greater with pure MA error processes than pure AR or ARMA error processes. The research can potentially benefit the applied researchers in Econometrics, providing empirical evidence that the FMOLS estimator performs relatively better than the OLS estimator for both cases. The results of the study may help the applied researchers choose the appropriate technique for analysis.   

Infrared Spectral Diagnostics: A Tool to Study Low-Density Massive-Star Magnetospheres

Program: Swenson Science Summer Research Fellowships
Faculty: Dr. Mary Oksala, Physics
Student: Alanna Walsh

Certain massive (~10 times the mass of the Sun) stars have large-scale, stable, strong magnetic fields. These stars also have very strong winds, which are known as stellar winds. The stellar winds are so strong that they blow particles from the star’s surface out into space. Because of the charged nature of the particles, they get stuck in the star’s magnetic field (or magnetosphere). When observing these massive stars in optical wavelengths, the star itself is much brighter than the often low-density, cooler materials in the magnetosphere. By observing using the infrared, the star dims, but the particles stay bright. According to theory, these particles can be trapped in the magnetic field because of the field’s strength, its stellar wind strength, and its rotational rate. To try to detect this material in cases where the optical fails to reveal the material, infrared spectral data was collected for several different known magnetic massive stars on two separate nights. A computer program called IRAF, or Image Reduction Analysis Facility, was paramount in analyzing the data. By using an observational approach called “nodding” along with several tasks within the IRAF program, much of the background noise was removed from the spectra. One of the most challenging issues with this data set was the removal of atmospheric features in the spectra from water vapor. A program called xtellcor_general in the IDL programming language did remove some of the atmospheric noise by comparing a target star to a corresponding standard star. At this point, the results are inconclusive. Although there were indications of the materials present in the data, there was still too much noise to positively and without a doubt confirm the materials’ presence. For a better chance at detecting the material, higher quality data may need to be acquired. 

Comparison of Mayfly Communities between Natural and Artificial River Habitats

Program: CLU STEM Research Abroad
Faculty: Dr. Mary Kelly-Quinn, Environmental Science
Student: Nathaniel Burola

Human beings have substantially modified rivers through the construction of road-crossings, dams, weirs, and other man-made barriers, with Ireland being no exception. This can have a severe impact on river hydromorphology and aquatic ecology. In some instances, man-made barriers can create large, artificial impoundments upstream of the structure which is characterized by low flow rates. As these impoundments can be extensive, it is necessary to understand the impacts they may have on river ecology. This study investigated the potential impacts of a ford and associated impounded reach on mayflies (Ephemeroptera) in an Irish river. Macroinvertebrate sampling was conducted in Brown’s Beck Brook in early May 2017. Kick samples were taken from three habitats: natural pools, riffles, and an artificial impoundment. Differences in species richness, total abundance, and abundance of individual species were compared between the three different habitats using PERMANOVA in Primer. No significant difference was observed between all three habitats regarding species richness. Total abundance of mayflies differed significantly across all three habitats, the abundance of mayflies in the impoundment samples was significantly lower than those from the riffle habitat. Baetis rhodani was the only mayfly species to show a significant difference in abundance between the three habitats, with the lowest number in the impoundment area. Certain mayfly species were unique or absent in certain habitat types depending on their adaptability to different flow conditions.  Rhithrogena semicolorata was absent from impoundment samples. In addition, Ecdyonurus sp. was absent from pool samples, and Paraleptophlebia sp. only occurred in the pool and the impoundment samples. Caenis rivulorum only occurred in impoundment samples. The potential effects of artificial impoundments on other species need to be further investigated in a range of river sites. This is being undertaken by the Irish Environmental Protection Agency-funded Reconnect project ( 

Effects of Legionella Pneumophila Type II Secretion on Temperature Dependent Growth

Faculty: Dr. Paloma Vargas, Biochemistry
Student: Savanah Gallegos

Legionella pneumophila (Lpn) is a gram-negative bacterium that naturally inhabits freshwater environments, but creates health concerns when it colonizes man-made water systems. This bacterium is the causative agent of Legionnaire’s Disease, a severe form of pneumonia that occurs after inhaling a mist of contaminated water droplets. Previous studies demonstrate that Legionella’s type II secretion system (T2S) is necessary for the infection of amoebae, which serve as natural hosts in the environment. Currently, individual mutant strains lacking expression for each of the twenty-five secreted proteins have been developed, but studies have focused on the role of T2S in host-infection at 37°C. The focus of the current study is to examine the effects of temperature on Legionella’s T2S mutant growth and infectivity. For all experiments, wild type Lpn (130b) and a type II mutant lacking a structural protein (LspF) were grown on solid media at 37°C for 72 hours before transfer to liquid media and experimental temperature conditions. Biological triplicates were created for each strain tested, and the optical densities were appropriately adjusted to ensure an equal number of colony forming units per sample. Strains were grown at 17°C, 25°C, or 37°C, and optical density readings were obtained to measure T2S mutant growth. Results indicate that the 130b wild type and the LspF mutant strain declined in growth after about 30 hours at 37°C. At 25°C, the LspF mutant declined in growth after 70 hours, whereas the wild type began to decline after about 100 hours, indicating a relationship between type II secretion and Lpn temperature-dependent growth. Further studies will explore the growth of LspF and additional mutant strains (such as LspDE and ProA) at various temperatures and in the presence of host amoebae to simulate environmental conditions.  

Biomechanics of Stair Ascent and Stair Descent in Middle-Aged Females and Males

Faculty: Dr. Michele LeBlanc, Exercise Science
Student: Gubidxa Gutierrez Seymour

Ascending and descending stairs are common activities in daily living. The ability to navigate stairs successfully affects a person’s safety in public settings and often determines if they can continue to live independently. Stair navigation is a complex task that becomes even more challenging when the aging process leads to physical deterioration, especially muscular strength, which begins as early as 25 years old. Stairway falls frequently occur, especially for females, across the adult lifespan and contribute widely to serious injuries. Stair research has been conducted on both young and elderly individuals; however little work has focused on middle-aged adults. Understanding the mechanics of stair ascent and descent for this population will better inform what strategies could be implemented to promote safe stair negotiation. The purpose of this study was to determine kinematic and kinetic values for middle-aged adults during stair ascent and descent. Subjects ascended and descended a short set of stairs at a self-selected pace at least four times with their left and right legs as lead leg, respectively. Lead leg order was randomized. Throughout each trial, eight Vantage 5 motion capture cameras and two Kistler force plates obtained 3-dimensional coordinates at 200 Hz and ground reaction forces at 1000 Hz, respectively. Knee muscle strength was determined by each subjects’ peak knee flexion and extension torques which were measured using isokinetic dynamometry. Subjects performed two warm up sets of five repetitions at 180°/second, then two testing sets of five repetitions at 60°/second, all of which were performed for each leg. The kinematic and kinetic analysis focused on ground reaction forces and hip, knee, and ankle joint angles and joint torques values. The results of this study will build upon prior biomechanical research on young and elderly populations to contribute to a clearer understanding of stair ascent and descent mechanics in healthy, middle-aged individuals.

Characterization and Analysis of Integrins on Immune Cells found in Canine Tumors

Faculty: Dr. Chad Barber, Biology
Student: Mary Jarjour

The purpose of this study is to identify and analyze integrins found on cytotoxic T cells, macrophages, and tumor cells present in canine tumors. Additionally, we aim to identify roles of these integrins in immune cells as well as other cells found in the tumor. Based on previous studies, I predict that the canine integrin aLb2 will be found on cytotoxic T cells and the integrin aMb2 will be found on macrophages in canine primary tumors analyzed. Primary methods for this study include cell culture and flow cytometry. Tumors received from veterinarians are rinsed with basal media and disaggregated; half the cells are seeded for cell culture. The remaining half is used for flow cytometry. We stained the cells with fluorescently-labeled antibodies that bind to specific integrins. The cells are run through the flow cytometer to detect integrin expression of the cells. We have been using antibodies both specific to canine species, but also some that are cross reactive for the canine. We have seen the expression of the aL subunit with two tumor samples. In previous studies, we have found expression of the a6 subunits on the canine cancer line, D17. We aim to perform inhibition assays with integrin blocking antibodies or peptides in vitro to assess the role in survival, migration or adhesion in the tumor microenvironment. By learning more about the integrins expressed on cytotoxic T cells and macrophages in canine tumors, we can make predictions about the role that the integrins play in cancer progression.