Poster Session 2

My project is the Rahman-Marshall Scholarship (RahmanMarshall.com), funded by the Marshall Fellowship. The Rahman-Marshall program provides high-school students in Pakistan with the opportunity to perform undergraduate-level research over the summer (guided by either a professor or a doctoral student). In Pakistan, opportunities for involvement in academic research at the secondary education (high-school) level are not available. Providing such opportunities encourages bright students to be exposed to academic research and develop a passion for innovative and rigorous academic thinking, which I believe is a vital component of education. We selected two promising, passionate high-school students into the program who successfully completed work this past summer with two scholars (one an F&M professor) in rigorous, challenging research. Information about the students is available on our website and their research presentations are available to view at the F&M research fair. This project was supported by funding from F&M's John Marshall Fellows Program.

Project Mentor: Professor Ed Novak, Department of Computer Science

Peer learning robots are an emerging educational resource that can be used to address the growing interest in computer science among students. Learning to code can be a tedious task and the ability for a peer learning robot to recognize frustration may improve the learning experience of the student. To demonstrate that a robot is capable of recognizing frustration, we built a frustration recognition component that uses facial action units to recognize when a user is frustrated in real time. The evaluation of the model's performance (F1 = 0.99) signified that the system is capable of accurately detecting when a user is frustrated significantly above random chance. The component was integrated into a larger architecture that uses eye gaze and speech to detect when a user interacting with the robot needs assistance. Future analysis will determine the effect of frustration recognition on the robots ability to recognize need. This project was supported by funding from F&M's Hackman Summer Scholar Program.

Project Mentor: Professor Willie Wilson, Department of Computer Science

Along with the introduction of English in schools in India came the spread of Britain’s "great literature." Shakespeare’s presence in India stems directly from this colonialist legacy. So why does Indian culture still feature so many references to and adaptations of William Shakespeare? Watching Bollywood films, listening to podcasts, and reading scholarly articles, my summer research sought to answer this question. Focusing mostly on Indian film versions of Shakespeare's plays, my project illuminated the differences between straight translations of a script and adaptations of a story. New questions arose, such as how changing the context, location, and words of a story impacts the original text. After centuries of Indian amendments to Shakespeare plays, can they still be considered Shakespearean? This project was supported by funding from F&M's Hackman Summer Scholar Program.

Project Mentor: Professor Padmini Mongia, Department of English

Cation integrates where the product lattice best matches the crystal structure of the starting material. The project determines the regioselectivity of consecutive cadmium and tellurium exchange. Cd exchange on roxbyite Cu2-xS nanorods integrates CdS along the c-axis, creating a top half region of CdS hexagonal. Performing Cd Exchange on Cu2-xTe/Cu2-xS heterostructure alters its regioselectivity, where CdS is integrated as stripes through the a-axis and shifting Cu2-xTe as stripes. Te exchange forms a shell of Cu2-xTe on Cu2-xS. The regioselectivity of Te exchange on CdS/Cu2-xS heterostructure does not change. At low incorporation of CdS, Te exchange drives CdS region as cores and forms Cu2-xTe shell, but Te exchange on high Cd incorporation cannot cause the CdS region to migrate. It is also seen that extreme Te exchanges breaks up hexagonal CdS lattice on CdS/Cu2-xS heterostructure, but Cd exchange at any level cannot break up hexagonal Cu2-xTe. This project was supported by funding from F&M's Hackman Summer Scholar Program.

Project Mentor: Professor Kate Plass, Department of Chemistry

The study of protein functions is important for the understanding of the cellular mechanism and biological pathways of organisms. Most proteins as we know only exhibit one function. However, there is a class of proteins named moonlighting proteins (MPs) in which a single protein performs multiple physiologically relevant biochemical or biophysical functions. Identification and study of these MPs through computational methods is crucial for research on diseases and drug-target discovery. In Summer 2022, based on 37 feature sets of protein sequences, we used Support Vector Machine (SVM), a machine learning algorithm, to resurrect a previously used method by others. However, no significant improvement in performance was observed. To achieve a higher accuracy, we then turned to using Position-Specific Scoring Matrices (PSSM) to establish new features for our proteins. We are now in the stage of experimenting with SVMs using our newly extracted features, hoping this will carry out improved results. This project was supported by funding from F&M's Hackman Summer Scholar Program.

Project Mentor: Professor Jing Hu, Department of Computer Science

In this talk, we will describe how we developed an automated image stacking pipeline for a long-term project optically monitoring a region in the constellation Orion that contains a small radio source (NVSS J054710-02319; hereafter, Source R). We will demonstrate how our process works, using some of 12,500 optical images obtained so far. Our image stacking pipeline uses a combination of routines from Astropy-affiliated packages (ccdproc, reproject, photutils) and Alsvid Python scripts. We first fixed bad columns in each image on our input list, then reprojected and cropped the images to center on Source R. Next we used mean subtraction to normalize the backgrounds, and used aperture photometry, which measures the light of specific stars, to further normalize the images, before fitting gaussians to those stars to do chi-square averaging, which weighs the poor seeing less heavily. This project was supported by funding from F&M's Hackman Summer Scholar Program and the Pennsylvania Space Grant Consortium.

Project Mentor: Professor Elizabeth Praton, Department of Physics and Astronomy

Plants often experience co-occurring biotic and abiotic stresses in their natural environment such as herbivory and drought or flooding conditions. The impact from flood or drought environmental conditions can impact the magnitude of attacks by leaf herbivores, such as Manduca sexta, and the variation in gene expression before and after herbivore attack in Solanum dulcamara. We examined the expression levels of three genes in the context of contrasting water regimes and differing herbivory damage using four genetically distinct maternal lines with 10 plants per each of the three water treatments and 15 receiving the herbivory treatment. We predicted that plants exposed to drought conditions would display higher levels of expression for genes PIN2 and comp572 and comp14877 would be downregulated relative to flooding and control conditions. We plan to use the sucking herbivore Lycorma delicatula to determine if differing feeding strategies cause differential gene expression in the Solanum dulcamara. This project was supported by funding from F&M's Hackman Summer Scholar Program.

Project Mentor: Professor Jorge Mena-Ali, Department of Biology

We report our findings regarding the long-term X-ray outburst evolution of the magnetar Swift J1818.0-1607 using The Neil Gehrels Swift Observatory. We used the data collected between the years 2020-2021. NASA’s HEASoft software allowed access to data visualization tools like Xselect and Xspec, which were used to represent the data thoroughly. The data were modeled using different spectral models, which provided information on the temperature, flux, and luminosity of the source. We found the X-ray flux to show a monotonic decrease over time, a consistent trend observed in other magnetars. The results from this project provide answers regarding the evolution of the spectral properties of the magnetars following an outburst. This project was supported by funding from F&M's John Marshall Fellows Program.

Project Mentor: Professor Fronefield Crawford, Department of Physics and Astronomy

In the Plass lab, experimentation of anion exchange with elemental selenium on copper sulfide nanorods in dodecanethiol altered the morphology in new ways we have never seen. This data contradicted previous anion exchange with tellurium that had successfully retained morphology and remained homogenous after full anion exchange. From this research, conditions were altered and enabled us to create a solid solution nanorod containing Cu/Se/S while retaining morphology. It was shown that as time increased, the incorporation of Se increased. This was of interest for the plasmonic and photocatalytic properties. This project was supported by funding from F&M's Committee on Grants Program and the Hackman Summer Scholar Program.

Project Mentor: Professor Kate Plass, Department of Chemistry

This multifaceted project aims to determine the efficacy of lead remediation in Lancaster, Pennsylvania while also investigating how the cognitive development of children between the ages of 2-11 living in these lead-remediated homes was impacted. In Lancaster City, over 200 homes have been remediated for lead since 2002, but the long-term effectiveness of lead removal is an area with very little research. Partnering with Property Lead Inspections, a certified lead contracting agency, data on the effectiveness of lead remediation was collected through survey responses and lead risk assessments. The cognitive development was assessed by administering the Minnesota Executive Function Scale and the Raven's Progressive Matrices test. Preliminary analysis shows that homes remediated more than 13 years ago are no longer lead-safe and that there is no significant difference in the cognitive scores of children living in lead-remediated homes compared to children living in control homes. This project was supported by funding from F&M's Hackman Summer Scholar Program, the Center for Sustained Engagement and the HUD Lead Research.

Project Mentor: Professor Jennifer Meyer, Department of Government and the Public Health Program

Trivalent carbons including organic radicals, anions, and cations are fairly reactive species yet have various applications in different fields of chemistry. We have designed several polycyclic aromatic hydrocarbon radicals, anions, and cations with a cyclopentadienyl central unit stabilized through resonance delocalization and steric protection. Organic radicals are utilized in Dynamic Nuclear Polarization (DNP) to enhance NMR signals; organic anions have applications as ligands in coordination chemistry; organic cations could help us understand the limit of antiaromaticity. Progress towards the syntheses of anthracene substituted tetrabenzofluorene/phencyclone radicals (3b/4b) and C55H55 “superbowl” radical (16·), anion (16-), and cation (16+) will be presented. Multiple synthetic routes were developed for each target and DFT calculations were performed to understand spin/charge distribution of radical, anion, and cation. The radical precursors of 3b and 4b were successfully prepared in four and two steps, respectively, and in 30-31% overall yields. Future directions and hypothetical syntheses will be discussed. This project was supported by funding from F&M's Hackman Summer Scholar Program.

Project Mentor: Professor Edward Fenlon, Department of Chemistry

We introduce the notion of a Patch Sampling Schedule (PSS), that varies the number of Vision Transformer (ViT) patches used per batch during training. Since all patches are not equally important for most vision objectives (e.g., classification), we argue that less important patches can be used in fewer training iterations, leading to shorter training time with minimal impact on performance. Additionally, we observe that training with a PSS makes a ViT more robust to a wider patch sampling range during inference. This allows for a fine-grained, dynamic trade-off between throughput and accuracy during inference. We evaluate using PSSs on ViTs for ImageNet both trained from scratch and pre-trained using a reconstruction loss function. For the pre-trained model, we achieve a 0.26% reduction in classification accuracy for a 31% reduction in training time (from 25 to 17 hours) compared to using all patches each iteration. This project was supported by funding from F&M's Hackman Summer Scholar Program.

Project Mentor: Professor Bradley McDanel, Department of Computer Science

Almost every type of data in programming is stored in an object. An object can be as simple as a 32-bit number or an extremely complicated specially crafted data type. Every object in an app plays an important role. Our research aims to create a robust system that would help to extract these objects from closed source proprietary applications into a human-readable JSON format so they can be closely examined, giving the user a window into how objects operate within the app. Examining these objects as a JSON object has many possible applications: it can help in understanding how an app works by examining how the objects interact or what kind of information the app is collecting; it can also help a user export some object from an app and into another app. Another area of research that we are pursuing is finding other concrete applications of our technique. This project was supported by funding from F&M's Hackman Summer Scholar Program.

Project Mentor: Professor Ed Novak, Department of Computer Science

Following the Gold Rush of 1849, the sleepy town of San Francisco exploded into a dynamic and complex urban center. Starting with a meager population of fifty people in 1849, the population skyrocketed to about 56,000 by 1860. While, early on, this growing population was overwhelmingly male, over time, single and married women came to the city in search of opportunity. In my research, I am examining how newly arrived women made their mark on an ever-growing urban landscape. Through my visits to several archives in San Francisco, I have recognized that by shaping and developing the urban landscape, these women are asserting their agency in an otherwise male-dominated world. Through civic reform, construction, and campaigning, women in San Francisco took advantage of the city’s ever-developing nature to gain a form of power as well as shape the city in their desired image to meet higher moral standards. This project was supported by funding from F&M's Committee on Grants Program.

Project Mentor: Professor Abby Schrader, Department of History

Proto-planetary nebulae exist as a small moment in a star's lifetime. There aare only a few to study. Using data from 30-meter and 12-meter telescopes, we are able to study these objects and their chemical composition. Protoplanetary nebulae have a strange abundance of sulfer isotopes when compared to standard sulfur-based molecules. This project was supported by funding from F&M's Hackman Summer Scholar Program and the Faculty Program Development Fund.

Project Mentor: Professor Deborah Schmidt, Department of Physics and Astronomy

Program decompilation is widely used for security analysis, code traceability, and bug fixing for software systems where source code is unavailable. However, modern decompilation tools take years to develop based on complex decompilation rules based on the target language. Additionally, it is often difficult to restore the information in the source code (such as variable names, functional information with specific meaning, and comments) as written by the author. In this work, we propose a new deep learning neural decompiler that converts Java Class files into accurate, user-friendly Java files. This approach relies only on (source code, bytecode) pairs and does not require any domain knowledge of the target language. We find that our approach generates more human-readable output compared to software-based decompilers while achieving a reasonably good compiler pass rate. This project was supported by funding from F&M's Hackman Summer Scholar Program.

Project Mentor: Professor Bradley McDanel, Department of Computer Science

The purpose of my research was to synthesize a copper complex that can function as a less expensive substitute to platinum as an effective O2 reduction catalyst. Fuel cells rely on chemical reactions to generate electrical power from chemical fuels. In the most common reaction coupled to fuel oxidation, oxygen is reduced and water is produced; however this process is kinetically slow so a catalyst such as platinum must be used. Platinum is very expensive however, so the need for an alternative catalyst is apparent. Towards achieving this goal, the synthesis of a tripodal ligand was optimized, adding copper to the ligand with different copper reagents, and conducting studies on the resulting copper complexes in an effort to learn more about how they behave. Studies including stoichiometry experiments, electrochemical studies, UV-Vis studies which consisted of reacting the complex with oxygen, and studies on analogous zinc complexes will be discussed. This project was supported by funding from F&M's Hackman Summer Scholar Program.

Project Mentor: Professor Davide Lionetti, Department of Chemistry

Access to postpartum mental health resources is vital to the healthy development of a mother and their newborn. This is especially necessary for mothers who do not have an effective support system at home and don’t have the economic resources to access postpartum care. A review of the Lancaster Community Needs Assessment and several interviews with community experts indicated that in Lancaster there is an increased need for OnDemand mental health support services for mothers. Especially following the Covid pandemic, there is a higher demand for free and reliable postpartum supportive services that can be accessed from home at any time. To address this we developed a prototype of a multifunctional online platform for postpartum mothers that can be easily accessed and provides reliable supportive services. The services include a video call hotline service, a chatroom for mothers to communicate with each other, as well as informational videos and podcasts. This project was supported by funding from F&M's Creativity, Innovation and Future of Work Program.

Project Mentor: Professor Maya Greenshpan, Department of Italian Studies and Hebrew

Sulfur-containing compounds have unique properties which can be used to develop new reactions in organic chemistry. Our project focuses on S-aryl thioformates, which are an aromatic sulfur-containing compound. Previously, we discovered that these thioformates produced carbon monoxide gas when reacted with a base. This summer, we investigated a new reaction of S-aryl thioformates with an alkene in a thiol-ene process to form a new addition product. In the literature, similar reactions are often done with more expensive catalysts such as gold; however, we were able to produce reactivity utilizing less expensive materials. Consequently, our summer research focused on understanding the mechanism of our reaction as well as optimizing the reaction to produce more desirable yields. We are currently working trying to determine the scope of the reaction by testing the effects of different functional groups. This project was supported by funding from F&M's Hackman Summer Scholar Program and the American Chemical Society.

Project Mentor: Professor Sarah Tasker, Department of Chemistry

Lyman-alpha emission is commonly produced by galaxies, and several cosmological surveys are trying to measure the structure of the Universe using galaxies detected in this emission. These surveys generally measure light from the inner parts of galaxies, but previous work has shown that much of the Lyman-alpha emission comes from an extended "halo" around the galaxy. We are using Python to simulate thousands of galaxies and understand whether surveys will detect their Lyman-alpha emission. We generate simulated galaxy images with a central component, a halo, and blurring by Earth's own atmosphere. We then calculate the fraction of the total Lyman-alpha light that would be observed by various planned surveys. Our initial results indicate that the fraction of the total Lyman-alpha emission that can be measured depends strongly on the fraction of light in the halo component vs. the central component of the galaxy, with weaker dependence on other properties. This project was supported by funding from F&M's Hackman Summer Scholar Program and the Pittsburgh Foundation.

Project Mentor: Professor Ryan Trainor, Department of Physics and Astronomy

RNA has been catapulted into the limelight of pop-culture conversation, as the powerful molecule plays a key role in Covid-19 vaccinations. This project seeks to develop a new tool to study the molecular dynamics of RNA. Through the use of theoretical coupling models, the distance between vibrational probe pairs may be extracted after measuring their vibrational coupling through 2D IR. This will allow the study of RNA folding and unfolding events. In order to develop this technique, nucleosides synthesized at F&M will be incorporated into an RNA hairpin by the Micura group in Austria and studied by 2D IR by the Tucker group at the U. of Nevada, Reno. Progress towards synthesizing nucleosides containing an isotopically middle-15N-labeled azide (N15NN) will be presented. The process by which the middle-labeled azide anion was successfully incorporated into the 5-methyl position of thymidine with an 8% yield and future directions will be discussed. This project was supported by funding from F&M's Eyler Fund and the National Science Foundation.

Project Mentor: Professor Edward Fenlon, Department of Chemistry

Analogical reasoning is the phenomenon of attending to the relational properties of a pair of items and applying it to a second pair. The ability to carry out this process is important as it facilitates learning, deepens conceptual understanding and generates novel ideas (Gentner, 1988). Emerging evidence suggests that children as young as three-and-four-year-olds can reason analogically in the presence of certain cues, e.g., linguistic labeling, comparative examples and causal cues (Goddu et al., 2020). In the present study, we explore whether social cues, which significantly enhance children's learning and memory (Howard et al., 2019), would help this age group reason analogically. We compare infants across four groups (Causal-Social, Causal-Non-Social, Non-Causal-Social and Non-Social-Non-Causal) to determine which cues best lead to the presence of analogical reasoning. Due to our small sample size in each group thus far, results are inconclusive and not trending in any one direction. This project was supported by funding from F&M's Hackman Summer Scholars Program.

Project Mentor: Professor Lauren Howard, Department of Psychology

Molecular probes are used to study protein folding, drug binding, RNA folding, and other structural and dynamic features of biomolecules. This project utilizes vibrational reporters– a newer type of molecular probe– to determine distances in nucleosides by measuring the vibrational coupling between them via 2D infrared (IR) spectroscopy. This will be the first major work to observe the coupling between two different vibrational reporters, an azide and a nitrile, on a hydrogen-bonded heterodimer that imitates a Watson-Crick base pair. The synthetic pathway to produce pyridines with a vibrational reporter at the 4-position will be presented. Di-N-butyl (4-azidopyridine-2,6-diyl) dicarbamate and di-N-butyl (4-(cyanophenyl)pyridine-2,6-diyl) dicarbamate were synthesized in 5 steps with moderate yields (11% and 42%, respectively). An azide peak at 2117 cm-1 and a nitrile peak at 2227 cm-1 in the IR spectra, as well the 1H NMR and MS, supported the successful synthesis of each. Future plans for alternative syntheses and IR experiments will be shared. This project was supported by funding from F&M’s Eyler Fund and the National Institute of Health.

Project Mentors: Professors Edward Fenlon and Scott Brewer, Department of Chemistry

The purpose of this project was to examine and synthesize archaeological research on artifacts uncovered from the sanctuary of Poggio Colla to discover patterns of votive behavior conducted by ancient Etruscans. The archaeological site of Poggio Colla, located approximately 20 kilometers NW of Florence, Italy, is a unique example of a sanctuary preserved within a wider historic and geographic context. Over 20 years of excavation unearthed large quantities of materials involved in votive behavior. My task this summer was to study various accounts of 12 distinct votive deposits for patterns in their history, purpose in ritual, and significance for the inhabitants of Poggio Colla. This synthesis unveiled themes prevalent to understanding not only the specific objects, but the importance of reverence and ritual for the civilization. This project was supported by F&M’s Hackman Summer Scholars Program.

Project Mentor: Professor Gretchen Meyers, Department of Classics

We present results for UV-selected active galactic nucleus (AGN) candidates, selected through the KBSS spectroscopic survey for star-forming galaxies at z~2-2.5. The classification scheme of N2/S2 Baldwin, Phillips & Terlevich(BPT), and Mass-excitation diagrams(MEX) diagrams were applied, which differentiated between star-forming and AGN types. These diagrams compare emission line ratios from different galaxies, and the ones that display strong ionization strengths are typically identified as an AGN. Additionally, differentiation between low-ionization nuclear emission-line regions(LINERs), and strong-ionization nuclear emission-line regions(Seyfert) AGN-host types were drawn in the S2-BPT diagram. The rest frame UV composite spectrum is created via stacking, and displays several strong ionization emission lines characteristic to AGN activity. These results reflect a comparison at high-redshift between star-forming galaxies and that of which hosts AGN activity. This project was supported by funding from F&M’s Hackman Summer Scholar Program.

Project Mentor: Professor Ryan Trainor, Department of Physics and Astronomy

Social assistive robots help people by providing verbal cues and gestures in response to social cues. Previous work has recorded 21 videos of a robot helping a person build a Lego creature. Now we extract data from the videos and investigate how the qualitative changes of eye gaze, head direction, and speech indicate the user’s behaviors and could enable the robot to predict when to provide help. Data indicates that when the head directions are directed at Lego pieces and no speech detected, there is a high likelihood of working, and the person does not need assistance. When eye gazes are directed at a picture of Lego, there is a high likelihood of thinking, and the person may need assistance. These results enable us to construct better models for the robot to recognize when to provide timely assistance, then people may have a better experience and feel more supported. This project was supported by funding from F&M's Program Development Fund.

Project Mentor: Professor Willie Wilson, Department of Computer Science