Natural Sciences

In this talk, a class numerical technique of finite difference method is presented for the solution of second-order one dimensional linear inhomogeneous damped wave equation. The stability and consistency of the method are discussed by using the stability matrix analysis. The numerical solutions of some damped wave partial differential equation models are presented. The results obtained are compared to the exact solutions as well as ordinary explicit and implicit finite difference method. The general idea of these methods is developed by using the C0-semigroup operator theory.

Previous research into the effect of height on perceptions of authority shows taller people have greater perceived leadership capabilities in various hierarchical settings. In our study, we aimed to examine whether this effect is also found on the reality television show Survivor, where jury members vote in a final tribal council for a winner. Data collection consisted of screenshots from every season’s final tribal council to analyze perceived height differences between season finalists when seated on stools of varying heights in front of the jury. We found a significant effect of perceived height on winning, even when accounting for factors such as sex and 2 versus 3-person finalist seasons. Our data suggest the finalist with the highest combined stool-contestant height has a higher chance of winning Survivor, and opens the door to further discussion on the impact of biases in our everyday lives.

Just as positive integers can always be factored into their prime components, elements in many other semigroups can also be factored into irreducible elements or atoms. The classification of these irreducible elements and the study of how elements factor as a product of atoms is a branch of mathematics called Factorization Theory. Historically, the focus has been on factorization in commutative semigroups, and only recently has an investigation of noncommutative settings been undertaken. We study the noncommutative semigroup of 2x2 matrices where all entries are nonnegative integers. Here factorization is highly non-unique. We show this by computing several numerical invariants. In particular, we identify irreducible elements and units and classify "almost prime-like" and "absolutely irreducible atoms."

Recent research has found a direct relationship between the amplification of the protein, N-MYC, and the aggressiveness of certain cancers, including neuroblastoma(Rickman et al, 2018). Additional studies have demonstrated that the protein, USP22, has a significant influence on N-MYC's close cousin, C-MYC, in that USP22 unnecessarily removes signals on C-MYC that would normally mark it for destruction. Without this normalized degradation, cancer progression can ensue.

To investigate the relationship between USP22 and neuroblastoma proliferation, we upregulated and depleted USP22 levels in vitro. The cells were then analyzed using rna sequencing, followed by meta-analysis using a pandas program developed in python. This approach allowed for an unbiased assessment of the genes differentially expressed following changes in USP22.

Our analyses demonstrated that following USP22 upregulation, N-MYC and other oncogenes were also upregulated. Further research will focus on upregulating and depleting genes most affected by USP22 alteration to investigate the pathways involved in neuroblastoma progression.

Muonium, the electron and muon bound state, is a quantum state of two particles subject to a Coulomb potential. Calculation of two particle bound state energy levels provides a specific prediction for spectroscopy experiments. We present an analytic calculation of the correction to the electron and muon bound state energy levels due to two photon exchange. In our calculations, we use Feynman diagrams and Feynman rules to set up the integrals. We also use Mathematica and the program FeynCalc to take traces in d-dimensions and expand the result of integrals through Laurent series to isolate the divergent contributions. Our result combined with other contributions gives the complete, finite, energy level correction.

My research project investigates a new analytic understanding of how rats initially learn to take drugs. By using a drug self-administration procedure, in which rats learn to lever press in order to receive cocaine infusions, I will assess different characteristics of that learning process. For example, how rapidly do animals learn to press the lever? It may only take some animals two days to learn but in some it may take six days. Or are there differences in the total amount of drug the animals take once they do learn the response? I will use a 3-parameter sigmoidal curve function to quantify any differences in learning to create “profile” of future drug abuse liability. This project will analyze biological sex and drug dose as independent variables. It is hypothesized that females and animals with access to the higher drug dose will have initiation profiles indicative of greater drug abuse liability.

CNTNAP2, a protein coding gene on chromosome 7, helps develop frontotemporal-subcortical circuits critical for human executive function precursors. Lacking CNTNAP2 within these circuits contributes to disruption of neurodevelopmental processes underlying various neurodevelopmental disorders, such as Autism Spectrum Disorder (ASD). Here, we describe behavioral phenotyping of the CNTNAP2 deletion line. Knockout (KO) mice have weight reduction in early adolescence and significantly weakened fore paw strength in adulthood, but no difference in hind paw strength, regardless of sex. Male and female KO mice perform better in a high-speed rotarod KO males and females show increased cold, but not heat, sensitivity in adulthood. KO animals have decreased input/output startle response and struggle with habituation in adulthood. No significant differences in sensory motor gating, anxiety-like behavior, social preference, or fear conditioning were noted. Our study provides a more thorough phenotypic profile of the CNTNAP2 deletion in mice, taking development and sex into consideration.

Factorization of whole numbers as products of primes is always unique. For example, the number 12 can only be factored into primes (or atoms) as 2x2x3. However, not all semigroups have this property of unique factorization and it is sometimes even possible that a product of two atoms can also factor out as a product of 4 atoms. In the semigroup of n x n upper triangular matrices with nonnegative integer entries, not all atoms have the “prime” property of whole numbers that if an atom divides a product, it must divide one element in that product. Instead, in this semigroup of matrices, for each number between 1 and n, there is an atom that divides a product of that many atoms and no subproduct. We measure this phenomenon by computing the so-called omega invariant for each atom, showing how different factorization is in this semigroup.

Permafrost, or perennially frozen ground, is found adjacent to cold, glaciated landscapes in many regions today, and is associated with periglacial environments. One notable periglacial landform is lobate to bench-like shapes of frost-shattered sediment, formed as sediment creeps and flows downslope due to repeated cycles of freezing and thawing in the uppermost layer of permafrost. The remains of these features are visible through LiDAR (Light Detection and Ranging) topographic data. Using LiDAR, relict lobate features can be mapped throughout Pennsylvania south of the Last Glacial Maximum ice margin. This study focuses on inactive lobes along Blue Mountain, the southernmost ridge of the Appalachians, and utilizes ArcGIS software to map and determine the geometry of these landforms. Their geometry is very similar to present-day active gelifluction lobes in Norway and elsewhere. In addition to landform analysis, an interactive and visually engaging presentation on Pennsylvania’s periglacial landforms was made using ArcGIS StoryMaps.

Dynamic Information Flow Tracking (DIFT) is a popular technique used to make various types of software more secure and is used extensively in Android security research. We propose a novel approach to implement DIFT in Android apps by modifying their bytecode. To do this, we implement our own system called "Stigma". Stigma re-writes an app by first converting its bytecode to Smali (a more human-readable variant of the dex bytecode). Our system then parses the Smali code of an app and adds DIFT source, propagation, and sink logic to the code. Our system repackages the app to be used as it normally would. By making Stigma robust and efficient, we create a sandbox that can be used to experiment with different approaches to solving existing problems in DIFT including but not limited to: implicit flows, tainting collections, and end-user security configuration.

The "metallicity" of a galaxy is the fraction of its elemental composition that is heavier than helium, and it plays a crucial role in the chemical evolution of galaxies. Measurements from the early Universe have recently become possible, and they would complement existing models. This project aims to determine the metallicity of a group of faint galaxies at z~2.5 with the help of Keck Observatory located in Hawaii. Previous attempts using the [NII 6585Å]/Hα line ratio method with a 1D-stacked spectrum yielded inconclusive results, and this method is suspected to be unfit for the early Universe. We employed a new 2D stacking method in order to minimize information loss, and a 26-sample bootstrapping analysis yielded a metallicity measurement of 12+log[O/H]=8.25±0.12, which is 0.36±0.10 of our Sun's metallicity. Our result will likely improve with a larger sample, and we hope to test its consistency with other methods in the future.

The Q1603 Lyman-alpha blob is a galaxy at a redshift of 2.55, corresponding to a distance of 11 billion light years away. Images from the Keck Observatory in Hawaiʻi show very strong emission at the Lyman-alpha wavelength (1216 Å) about 100 billion times more luminous than the Sun, spanning a diameter of about 600,000 light years. We measured the emission line ratios of this galaxy for hydrogen, nitrogen, and oxygen, known as the N2-BPT line ratios, to determine the source of the strong Lyα emission that we observe. Comparing these values to measurements of Sloan Digital Sky Survey galaxies, we conclude that the Q1603 Lyman-alpha Blob is powered by a supermassive black hole that is blowing out high-energy material and ionizing the gas within the galaxy.

The goal of this project is to create a swarm of robots to sweep a minefield to mitigate the risk of human harm. This is done by having a lead robot go out into the field with an impulse radar. If it detects a possible mine, then it will drop a virtual waypoint on top of the location and then continue on. The waypoints are then sent to a shared database and create a path. Once the path is set, two other robots are then sent out on the path and rescan the area with an imaging metal detector and a ground penetrating radar respectively. All data is then sent to a shared database and combined into one image that will be used to determine if the object is actually a landmine. My role in this research is to test the scanners, combine the different scans, and now help with getting the robots to follow the virtual path.

The main goal of our research is to measure how non-unique the factorization can be in the semigroup of upper triangular matrices over nonnegative integers. After defining this semigroup, we recall which elements are atoms, or irreducibles. We then define union of sets of lengths which is one tool that can help us measure how non-unique factorizations can be. We share the results of our calculations which compute the union of sets of lengths containing k for n x n upper-triangular matrices for all positive k and n.

For my summer 2020 project, I redeveloped parts of Longenecker et al.’s software tool (2017), ECHO-GPM, that will allow our research group to uniquely calculate the quantity, location, and correlation of groundwater recharge and discharge areas. The redeveloped code will cross-correlate user-collected data of the pressure in a groundwater spring with rainfall data from NASA’s Global Precipitation Monitoring (GPM) satellites. This involves analyzing the Python algorithm, reassessing prior ECHO-GPM results, and filtering suggestions from our research group. The primary objectives are to incorporate different cross-correlation methods and reduce the computational cost of running the code, considering the large datasets. This means exploring cost-reducing options, such as multi-parallel processing for the cross-correlation performance. These calculations, previously done by hand, can now be automated to inform potential solutions for groundwater exploration, including issues of water scarcity due to global warming and the depletion of surface and subsurface water from over abstraction.

Testing is crucial in education as it reveals students’ knowledge about a certain topic. However, whether conventional testing methods such as multiple choice is the best method to reflect students’ true ability remains unknown. The purpose of this study is to compare multiple choice with probabilistic scoring and determine which testing method is better. Unlike most previous research, which compares these two methods qualitatively, we use classical test theory and item response theory to generate quantitative data for comparison. We asked participants to complete the Mechanics Baseline and Force Concept Inventory tests using either multiple choice or probabilistic scoring. Data collected were smoothed using a Gaussian smoother to reduce noise so that metrics from Item Response theory could be used. This project was made possible by funding from F&M's Hackman Summer Scholars Program.

Down Syndrome (DS), a genetic disorder due to an extra copy of chromosome 21, leads to mortality and immunological alterations. We studied the Dp(16)1Yey mouse model gene expression in liver using RT-qPCR. Probing the expression patterns of triplicated genes during immune system development may indicate the origins of immunological abnormalities and blood stem cell imbalances in DS. We compared transmission rate of the Dp(16)1Yey chromosome to offspring surviving at weaning based on the sex of the carrier parent for 143 litters. The mean rate of Dp+ mice at weaning was 26% with both maternal and paternal carriers, indicating no significant difference, and refuting the theory that maternal phenotypes affect transmission rate and survival of offspring. A post-weaning death analysis found significantly higher juvenile mortality in mice with the chromosome duplication (Dp+) as compared to euploid mice. These findings indicate the DS mouse model replicates some phenotypes of the human disease.

Neuroblastoma is among the most common childhood tumors, and previous research suggests the influence of N-MYC and USP22 proteins on tumor progression. This project explored the hypothesis that in neuroblastoma, the protein USP22 is involved in deubiquitination that prevents N-MYC protein degradation. Previous research has demonstrated that amplification of the N-MYC protein is associated with neuroblastoma progression and that USP22 influences the activity of C-MYC, a protein related to N-MYC. In this project, endogenous expression levels of NYMC and USP22 in various cell lines were characterized and the effect of N-MYC and/or USP22 overexpression on transcription of various genes was explored with qPCR. Multiple methods of co-immunoprecipitation were attempted to investigate the formation of a complex between USP22 and N-MYC. No clear relationship between N-MYC and USP22 protein or mRNA was determined in this project, and future work should continue to explore the potential relationship between NMYC and USP22.

ArcGIS StoryMaps are powerful web-based applications that combine interactive maps, text, images, and videos into a multi-media, map-centered, user-controlled narrative website. They provide a compelling, effective way to summarize complex geological and environmental issues to a wide audience. The goal of my project is to develop a StoryMap that effectively explains the objectives and results of a long term restoration experiment at Big Spring Run (BSR) in Lancaster County, PA. This multi-year BSR valley bottom restoration experiment tests the efficacy of a novel design to improve aquatic and riparian ecosystems, and to reduce nutrient and sediment loads, among other objectives. The completed StoryMap will be made available to the public and will be used by policymakers and restoration practitioners to target other streams for effective restoration.

My summer internship with Regenerate Lancaster involved researching, writing, and publishing solutions that will reduce carbon emissions in Lancaster County and benefit the economy and community. This served as a valuable extension of Professor Hirsch’s Environmental Migration Lab as it focused on how to make immigrants' new home and community more sustainable as efficiently, safely and equitably as possible. Two solutions I found particularly interesting are Tree Intercropping and Low-Flow Fixtures. Tree Intercropping integrates perennial tree crops into areas of annual cropping, which increases farmers’ profits, makes them more resilient, and positively impacts their land by increasing the carbon content and productivity of the soil, and reduces erosion. Low-Flow Fixtures for showers, taps, and low-flow toilets significantly reduce household water and energy usage, and reduce greenhouse gas emissions.

This project will contribute to a better understanding of the current environmental conditions of watersheds in Lancaster County and place this understanding in a historical context. The natural landscape constantly changes over various timescales in response to a multitude of factors. Natural factors over geological timescales have been superseded by human impacts, especially over the last several decades. Fishing Creek in southern Lancaster County is one of the least impacted watersheds and includes areas of land preserved by the Lancaster Conservancy. My goal was to build a GIS database including a comprehensive set of georeferenced aerial photographs from the 1940’s to the present. This dataset was then used to identify areas where the stream channel, and the associated riparian buffer zone, have changed over the last 80 years. This information will be used to better understand the dynamics of the stream corridor and target areas for conservation in the future.

Our world is experiencing an unprecedented crash in biodiversity, in part because of urbanization. Urbanization increases human-animal interactions, leading to only a subset of species being able to thrive in a novel, fragmented, and rapidly changing environment. Under the guidance of Daniel Ardia PhD., our lab investigates the factors affecting the mammal diversity in remnant forest fragments.

Building on this work, my proposed senior research project will look at how urbanization and ecosystem structure influences the distribution and abundance of raccoons, a common reservoir for rabies. I will use these patterns to make predictions about human rabies infections within southeastern PA. Despite the availability of a vaccine, rabies continues to be a public health problem that affects human society. Understanding the ecology of rabies within southeastern PA will help government and conservation agencies develop policies that can prevent further spillover events of this deadly viral disease.