Schedule of Events

The Drosophila melanogaster compound eye consists of approximately 750 single eyes (ommatidia). Each eye is separated horizontally by a Dorsal-Ventral (D-V) midline, an “equator,” with ommatidia dorsal to the equator having their photoreceptor cells arranged in a trapezoidal pattern that is a mirror image to those of ommatidia ventral to the equator. During development, the equatorial cells are an important signaling center for proper eye formation. Flies were obtained from Dr. Catherine McElwain (LMU) that incorporate an insertion of a white+ (w+) transgene into an as yet unmapped location on the third chromosome. The phenotype of this line is novel in that the w+ transgene’s expression pattern, visible in a w– genetic background and unlike most w+ transgene expressions, consists of low-level w+ (pigmented/wild-type) expression throughout the eye, with a much stronger expression of the transgene in the equatorial region, an expression named “Racing Stripe.” We have obtained variations of this expression pattern, modified in different genetic backgrounds (Pin Stripe, Wide Racing Stripe, Scythe). This variability is suggestive of epigenetic modification that alters the w+ transgene’s expression. We have investigated whether its expression can be affected by mutations known to alter heterochromatin status. Crosses show that changes in heterochromatin status can alter the expression of the w+ transgene. We have mapped the genomic region of this transgene to investigate its possible relationship with unidentified genes that may also be expressed along the D-V midline. This work holds implications for gene regulation and genome engineering.

In 2012, using the largest particle collider in the world at CERN in Switzerland (the LHC), physicists found the Higgs Boson, a particle which had been hypothesized but had eluded detection until then. A reason that this particle is important is because the Higgs Boson is linked to why some particles have physical mass. There are four main processes that can produce the Higgs through a proton collision (the type of collisions being observed at the LHC), and the only one that has not yet been experimentally observed is called Vector Boson Fusion (VBF). The focus of our research is to produce a machine learning algorithm that can sort some of the LHC collision data: accepting VBF Higgs while rejecting background processes. In my presentation, I will show the effect of our selection cuts in increasing the population of expected Higgs events while decreasing the population of ZZ events in our filtered data. A machine learning neural network was then trained on these filtered data and further increase the ratio of Higgs to ZZ events in our sample. I am presenting the work on our attempt to optimize the machine learning algorithm to more reliably select Higgs from the background through a variety of measures, including: smoothing the input files’ (Ntuples) kinematical variable distributions, experimenting with the structure of the neural network (manipulating the number of hidden layers), and excluding different kinematical variables from our network’s inputs.

Foraminifera (single-celled marine organisms) are one of the most important paleoclimate tools because their tests, or shells can be used to reconstruct past ocean temperature and sea levels. Foraminifera build their tests in equilibrium with the surrounding ocean water and provide a snapshot of the ocean chemistry during their lifespan. The oxygen isotope ratios (δ18O) measured in the tests of foraminifera provide ocean temperature information. Higher δ18O values represent cold glacial periods and lower δ18O values indicate warm interglacial periods. When reconstructing past climates, it is important to use well-preserved foraminifera. When a poorly-preserved foraminifera is used, the δ18O signature can be offset leading to a skewed paleotemperature and/or paleo sea level. Previous research by Poirier et al. (2021) quantifies the effects of different levels of test preservation (glassy, pseudo-glassy, frosty) on past climate records. In our study we expand on this work by increasing the sampling resolution at Site 846 (Pacific Ocean) spanning ~115,000-165,000 years ago. Marine core samples are prepared using standard sediment techniques and all specimens of Cibicidoides in a sample are removed and categorized by the level of preservation and breakage to determine the Foraminifera Preservation Index (FPI). Our results show that as we move from deeper (older) samples to shallower (younger) samples, there is a gradual increase in the preservation score. This trend is likely due to a more corrosive water mass in the Pacific ocean during interglacial cycles. This work is important for paleoclimatologists who utilize foraminifera δ18O in climate modeling.

The ability of Desferrioxamine D (DesD) to catalyze the sequential formation of non-ribosomal peptide bonds is unique and understudied. Interestingly, DesD catalyzes multiple amide bond formations on the monomeric substrate N-hydroxy, N-succinyl cadaverine (HSC) to give a dimer, a trimer Desferrioxamine G (dfoG), and ultimately the macrocyclic siderophore known as Desferrioxamine E (dfoE). Uncovering the biophysical details of this synthesis remains of interest given that the target molecule is a clinically useful iron-chelating agent and that inhibition of DesD is a possible strategy for combating pathogenic bacteria. 
In the absence of commercially available HSC, a Gabriel amine synthesis that was reported in the literature by others has been reproduced on a multi-gram scale. Regarding noteworthy modifications, we find that: 1) elevated H2pressure and Pd(OH)2 give a more reliable removal of the benzyl and carbamate protecting groups in the precursor to HSC, and 2) for final purification standard silica gel chromatography in absolute methanol is preferable to semi-preparative HPLC and gradient elution on a C-18 column. HSC prepared under these conditions gives colorless material that is analytically pure and of sufficient quantity for use in kinetic and thermodynamic assays. 
Access to synthetic HSC will facilitate the ITC analyses of all DesD variants with monomer as substrate and test our hypothesis of cooperativity and half occupancy as observed in binding studies. We will also explore the possibility of a two-step mechanism by analyzing reaction mixtures via HPLC after catalytic turnover to confirm the intermediacy of both dimer and trimer.

In order to understand any statement in a language, one must have a functional mental grammar of that language with which to parse it; this is true for natural and constructed languages. Toki Pona is a minimalistic constructed language currently consisting of about 137 words which was originally created by Sonja Lang in 2001. In 2012, Dr. Zach Tomaszewski of the University of Hawai’i Mānoa put forward a context-free grammar for the language in "A Formal Grammar of Toki Pona," but this grammar lacks structures and words now commonly used in online Toki Pona-speaking communities. Another grammatical analysis of Toki Pona that differs from Dr. Tomaszewski’s also exists, and a context-free grammar created from it might be more effective in parsing contemporary sentences and handling part-of-speech ambiguity, though that may come at the expense of efficiency. The ultimate goal of this research is to update Dr. Tomaszewski’s original grammar so as to be able to parse a wider variety of contemporary sentences, to make another grammar from an alternate analysis, and then to compare their performance on a set corpus using Tomaszewski’s original parser as a strong first step towards a more complete understanding of natural language processing.

Triathletes often complain about lower limb discomfort when running after cycling (Quigley, 1996). Several studies have found differences in muscle activation (Chapman, 2009), kinematics (Rendos et al., 2013), and kinetic cost (Millet, 2001) during the transition run. These differences were also found to be more severe in less experienced triathletes (Chapman, 2008). PURPOSE: This study aimed to determine the kinetic differences between baseline and transition runs of inexperienced triathletes. METHODS: Twelve novice triathletes age: 29.4±12.15 y, mass: 71.2±10.3 kg volunteered to participate. Each participant did one 20 minute control run and one run after a 20 minute cycling session. Running sessions occurred on an outdoor loop where subjects passed over two force plates (Kistler, 1200 Hz). A two-way repeated-measures ANOVA was used to determine differences between run type and time block (α=0.05). RESULTS: During the stance phase, there was a main effect for time block of anterior linear impulse (p=0.043) and average +RFY (main effect p=0.03) and contact time (p = 0.039). Similarly, there was a main effect for time of propulsive phase linear impulse (p<0.001) and average +RFY (p<0.001). Although post-hoc analysis did not reveal significant differences, it appears the later time blocks decreased anterior linear impulse as RFY decreased. CONCLUSIONS: Athletes displayed similar running kinetics between the baseline and transition run, however differences did occur between time blocks. This suggests that athletes fatigued similarly between the two run types. These findings may indicate that self-reported discomfort in the transition run may not be detected by measures of whole-body kinetics.