Project in Psychology and Neuroscience: “Neurological and Behavioral Responses to Urine and Other Volatile Odors: A Study of Pheromones“
Catherine Chen, Sakshi Chopra, Thomas Flynn, Anita Gosevska, Rohith Kariveda, Simran Kaur, Neetika Rastogi, Anna Rezk, Revika Singh, Audrey Yeung, Victor Zhang
Advisor: Dr.Graham Cousens
Assistant: Zachary Rissman
Pheromones are chemical substances released by mammals into the environment that may affect the behavior and psychology of other mammals of the same species (1). This study aims to investigate the role of the amygdala with respect to the accessory olfactory system as well as to understand the social transmission of fear via pheromones. To maximize the study’s implications, this experiment was split into two parts.The first part tested the physiological aspects of pheromonal exposure released by other rats in response to pheromonal, predator, and controlled stimuli; both innate and conditioned fear urine elicited stronger startle response than the control urine. It was concluded from that data that pheromones can be transmitted via rat urine, and thus are able to trigger fear responses in other rats. Overall, the experiment contributed to furthering the understanding of how fear can permeate through a group of rats solely through their sense of danger and/or nervousness from an individual rat. Given that these rat models maintain similar brain anatomy and physiology to those in humans, the findings can be applied to humans with regards to the role of the main and accessory olfactory system as well as the social transmission of fear through a set of individuals. The second part of the study utilized electrophysiology to understand the role of specific amygdaloid nucleus in the main and accessory olfactory systems, as well as to gain new levels of insight about the social transmission of fear in mammals and potential applications in humans. In the electrophysiology portion, rat models were exposed to volatile odors, conditioned fear rat urine, and naive rat urine and neurological activity was recorded. Based on a brain histology, the electrodes used for data collection reached the anterior cortical nucleus – part of the main olfactory system – rather than the medial nucleus – part of the accessory olfactory system – in the amygdala; the electrode location signifies reason behind volatile odor response and a lack of pheromone response. Had the electrode achieved the proper location, a more conclusive results could have been achieved.
Henry Chen, Vedant Dhopte, Elizabeth John, Joanna Kuo, Ashwin Mahadevan, Eli Malkovskiy, Shivani Patel, Sara Recarey
Advisor: Dr. David Cincotta
Assistant: Dominique Voso
Controlled release kinetics, the constant-rate diffusion of a substance, is a concept with many important applications in pharmacy, cosmetics, and agriculture. The purpose of this project was to gather data on the diffusion rates of various organic compounds through a polyethylene vinyl acetate (EVA) membrane and then model how their various properties affected the rate of pseudo-zero-order-release. Gravimetric analysis of membrane sealed jars that were partially filled with liquid compounds was carried out, and the change in mass over time due to vapor diffusion allowed for the calculation of vapor flux which was used to determine each compound’s diffusion coefficient. The models that were created related the compounds’ Hansen Solubility Parameters including strengths of London Dispersion forces, dipole-dipole attractions, and hydrogen bonding, as well as the compounds’ molar masses, and molar volumes, to the diffusion coefficients that were experimentally determined. Two separate models were created for hydrocarbons and polar substances because their properties were highly disparate. The polar substance model and the hydrocarbon model had R2 values of .348 and .476 respectively. The results of this project imply that there are multiple, complicated factors that are involved in a substance’s ability to diffuse through a certain membrane. However, the models created in this project are a base for what can eventually become an all inclusive and accurate predictive model for the diffusion rate of any substance through any membrane.
Lydia Chen, Alexis Fryc, Thomas Hontz, Christie Hung, Mofeyifoluwa Oluwalana, Anisha Shin, Shreyas Srinivasan, Armaan Tobaccowalla, Timothy Topolski, James Tsatsaros
Advisor: Dr. Minjoon Kouh
Assistant: Samuel E. Zorn
Through machine learning with the programming language Python, users can be identified by their unique features. “Machine learning” is a subfield of computer science in which computers are trained to process and learn from data without being explicitly programmed (1). Two programs were written to utilize machine learning with the goal of recognizing identifying characteristics of users. The first program matched an identity to a recording of a voice, a process called speaker recognition. 6 voice recordings from each of the 11 human participants, saying “Hello,” were gathered and preprocessed to eliminate white noise, normalize volume, and organize the audio recording into numerical data. Next, the data was classified using a multiclass classifier and analyzed by power over frequency graphs. Using the data training set, the program could predict the identity of an unknown participant’s recording. A confusion matrix illustrated that the program yielded 100% accuracy. The second program predicted a Twitter user’s gender based on a single random tweet from his or her account. Various users’ genders and tweets were compiled as input to the algorithm. Indiscriminatory characters from the dataset were cleaned, and two sets of predictors were assembled. The predictors were later inputted into the Multinomial Naive Bayes classification algorithm, which then outputted the highest probability gender for each tweet in the dataset, yielding 60% accuracy. Both programs showed an accuracy greater than or equal to that of humans.
Isra Arbab, Chinaza Asiegbu, Suraj Chandran, Amy Liu, David Oliveira, Kunj Padh, Anuj Patel, Manav Vakil, Elisha Zhao
Advisor: Dr. Ryan Hinrichs
Assistant: Zoe Coates Fuentes
Iron is an essential mineral for fundamental biological processes such as photosynthesis and nitrogen fixation. Oceanic bioavailability of dissolved iron regulates the primary productivity of phytoplankton, especially in high-nutrient low-chlorophyll (HNLC) areas. Such HNLC regions comprise one-third of the world oceans and are high in surface nutrients like nitrates and phosphates but lacks in iron. Iron in these regions originates primarily from mineral aerosols arising from dust storms in desert regions. This study investigated the effects of various atmospheric organic compounds found in atmospheric aerosol and cloud droplets, anthropogenic emissions, and solar radiation on iron solubility and bioavailability. Results indicated that some organic ligands (Suwannee River humic acid and catechol) had a positive effect on iron solubility in atmospheric waters by increasing Fe(II). Others had a neutral or inhibitory effect (Pahokee Peat humic acid, 2-nitrophenol, and isoprene). Both the Suwannee River humic acid (SHA) and catechol increased total iron solubility by over 50% and Fe(II) solubility by over 400%. Furthermore, SHA increased Fe(II) solubility from 0.181 μg Fe/mg ATD to 0.971 μg Fe/mg ATD, on average, and catechol increased Fe(II) solubility from 0.181 μg Fe/mg ATD to 1.21 μg Fe/mg ATD, on average. For the photoreduction study, photoreduction was found to have no net impact on iron dissolution as the data was inconsistent with respect to the control, suggesting the need for further research. The still lamp consistently inhibited total iron by approximately 75%, while the rotating lamp was inconsistent against the control. Based on the indeterminate results of SO2 on iron dissolution, as demonstrated by a consistently higher Fe(II) concentration (average of 0.102 μg Fe/mg ATD) than total Fe concentration (average of 0.0235 μg Fe/mg ATD), we found that further research is needed to correctly simulate the effect of SO2 treatment on Fe(III) reduction.
Anjali Gupta, Luke Jacob, Meena Mandalapu, Alexandra Poret, Aparna Ragupathi, Matthew Shih, Grace Simmons, Brian Song, Laura Tsai, Joy Xie, Sandra Yang, Victoria Zhou
Advisor: Dr. Stephen Dunaway
Assistant: Jal Trivedi
As fungal infection rapidly becomes a more imminent threat to the global health community, the need for a novel solution becomes increasingly necessary. Despite the potentially lethal effects of fungal disease, current treatment methods are largely ineffective and the search for a potential drug target to treat mycoses is crucial. Eukaryotic elongation factor 3 (eEF3), which plays an important role in protein synthesis, exists primarily in fungi and is inactive in complex multicellular organisms like plants and animals. This study evaluates the conservation of the eEF3 gene throughout lower level eukaryotes, particularly Coccomyxa subellipsoidea and Saccharomyces cerevisiae, in order to better identify eEF3 as a viable drug target. The eEF3 gene from the Coccomyxa subellipsoidea green algae was cloned, amplified by Escherichia coli bacteria, and then transfected into Saccharomyces cerevisiae. To test for gene complementation between the two species, this study employs the plasmid shuffle technique. Environmental pressure is placed on the endogenous eEF3 gene of the yeast in order to determine if the transfected C. subellipsoidea gene is an adequate replacement. Despite this selective pressure, the yeast cells failed to eject their eEF3 gene, suggesting that the eEF3 of the C. subellipsoidea did not functionally complement and could not substitute for the endogenous gene. This study provides valuable information regarding the limits of eEF3 as a potential drug target and the extent of its conservation among lower level eukaryotes.