Mason Journals (George Mason Univ.)
Not a member yet
3256 research outputs found
Sort by
Oaxaca, Mexico: Understanding and Teaching World/Global History Through the Study of an Early Modern Mexican City
No full textFederico De Romanis and Marco Maiuro, eds., Across the Ocean: Nine Essays on Indo-Mediterranean Trade
No full textProtein Signaling Profile of Prostate Tumor Metastasis Treated with Pembrolizumab: A Case Study
No full textThe immune system is heavily influenced by protein signaling pathways that drive the growth and metastasis of prostate tumors. Understanding the activation changes of immune-related proteins before, during, and after treatment can provide valuable information regarding treatment efficacy. This study aimed to investigate the protein signaling expression changes occurring within a patient’s tumor diagnosed with prostate cancer that metastasized into the bladder while being treated with Pembrolizumab. Nine biopsies were obtained at different times throughout the duration of treatment that were then used in our analysis. Using the instruments’ manufacturing recommendations, tumor cells were isolated from formalin fixed paraffin embedded (FFPE) tissue sections from the nine individual prostate cancer biopsies. In following lab procedures, these slides were deparaffinized, dehydrated, and hematoxylin stained before Laser Capture Microdissection (LCM). LCM caps were lysed using extraction buffer. The lysates that were obtained were then printed onto nitrocellulose slides using Reverse Phase Protein Array (RPPA) and stained with a group of antibodies selected to examine immune signaling and DNA damage repair endpoints. A laser scanner was used to scan the slides and the images produced were analyzed with MicroVigene software. Upon reviewing our results, biomarkers FoxP3, MSH2, CD3 zeta, and MLH1 showed elevated activation levels. In addition, Chk-1 (S345), Chk-2 (S33/35), STAT1 (Y701), PD-L1 (22-c-3), and PD-L1(E1L3N) appeared to be less expressed. Further research is needed on a larger sample size to support our results
Market Dynamics in Response to AI Breakthroughs: (Early) Evidence from ChatGPT
No full textMajor advancements in artificial intelligence (AI) have the potential to enhance productivity and generate wealth. This study investigates the stock returns of 20 leading AI-related companies surrounding three significant ChatGPT release dates: the initial release, GPT-4, and GPT-4o. Utilizing historical stock price data, we conduct event studies to assess the stock market responses. The analysis reveals substantial stock price changes associated with all three release dates. Notably, the GPT-4 release had the most pronounced effect, yielding a return of 6.66% and a t-stat of 5.98, signifying a strong impact on stock valuations. The initial release and GPT-4o also demonstrated significant changes, with returns of 3.94% and 4.55%, and t-stats of 3.07 and 4.47, respectively. These results indicate that major ChatGPT releases significantly influence the stock prices of AI-related companies, suggesting considerable financial market reactions to AI advancements
Exploring the Relationship Between the GitHub Development Activities of an OSS and its Funding/Governance Model
No full textGitHub is a collaborative tool used for the development of different software and projects. One type of software commonly sourced on GitHub is Open Source Software (OSS), a type of software in which the source code is publicly accessible and anyone can aid in the development of the project. This research aims to uncover whether there is a link between an OSS’s GitHub development activities like watch/commit ratios and its funding/governance models. Using BigQuery, we were able to collect the development activities of 500+ OSS projects through running a SQL (structured query language) script on GitHub Archive, a database of public GitHub records. After this, we collected key information on the funding and governance models of the OSS projects to determine if these attributes were correlated with the GitHub records collected earlier. Although the data has not been fully processed and analyzed yet, if a correlation is found it could reveal key insights into how to successfully manage and structure the development process for different types of OSS projects.
This abstract is part of a collection in which the overarching large project under Dr. Jiasun Li was subdivided into discrete critical tasks that were carried out by multiple individuals or smaller teams. Abstracts in this collection read similarly given the shared project goals, but represent distinct tasks completed by the abstract authors towards finalizing the described analysis
Analyzing the Relationship Between Project Success and Project Type for Open Source, Decentralized Finance and Web 3.0 Projects
No full textDecentralized applications and services are crucial to the crypto industry. This study examines the development history of over 600 open source crypto projects using GitHub Archive data from 2013 to 2023. By leveraging SQL scripts, this project extracted various project activities—such as watches, pull requests, pushes, commits, and branch creations—to build a comprehensive history for each project. Projects were categorized into types like crypto wallets, Dapps, L2 networks, tokens, and stablecoins to identify trends in GitHub activity by project type. They were also distinguished by their funding and governance models. Our hypothesis suggests that different project types may exhibit distinct activity patterns, such as Dapps having more commits or token projects having more branches. Additionally, our research focuses on the correlation between a project's governance and funding model and its success, measured by the level of user development and interaction on GitHub. It is hypothesized that more successful projects show higher and sustained activity levels. This analysis aims to determine if certain funding and governance models models, such as those of DAOs versus private companies, reliably correlate with project success, or if external factors play a larger role. The findings offer insights into the governance of decentralized corporations and the impact of business models on project success, enhancing our understanding of the factors driving development and engagement in the crypto and OSS communities.
This abstract is part of a collection in which the overarching large project under Dr. Jiasun Li was subdivided into discrete critical tasks that were carried out by multiple individuals or smaller teams. Abstracts in this collection read similarly given the shared project goals, but represent distinct tasks completed by the abstract authors towards finalizing the described analysis
Creating Complex Math Question Test Sets for LLMs to Train With
No full textLLMs, or Language Learning Models, (such as ChatGPT) can easily interpret human input and provide the desired output. However, when it comes to solving problems that cannot be conventionally analyzed, like math questions, many LLMs struggle to produce accurate results. In order to address this problem, we have created an algorithm that generates a set of complex math questions, which can be used as a training set in the machine learning process of LLMs. The algorithm asks the user for a number, and then generates and prints a dataset of that many questions. Each question is randomly chosen as an integral or derivative, with each component randomly generated (including nested functions) with 3 predetermined operators separating them. Due to Python’s lack of accuracy in generating answers, the questions are input into an advanced LLM with the capability to do calculus, and the output answers are stored in a csv file. The questions are then input into a less advanced LLM, and its output answers are likewise stored in order to compare it with the accepted answers. Using evaluation metrics in python, the accuracy score of the generated answers in one trial of 25 questions was 0.64, providing a baseline evaluation. Although current LLMs are very capable, many are unable to solve some complex problems that humans input. By using the described process, there exists the possibility to significantly improve the performance of LLMs by using these sets to train them to solve these problems and further aid humanity
Screening Aggregation Breakers of Alzheimer’s Disease Aβ Protofibrils Using AutoDock CrankPep
No full textAlzheimer’s disease (AD) is a neurodegenerative disorder putatively caused by the accumulation of β-amyloid (Aβ) peptides in the brain. To treat AD, drugs and small peptides that inhibit Aβ peptide aggregation are being explored as potential therapies. With the goal of elucidating the molecular interactions between Aβ peptides and potential inhibitors to explore their therapeutic effect, we used AutoDock CrankPep (ADCP) to simulate the docking of small peptide inhibitors that act as aggregation breakers preventing incoming Aβ peptides from adsorbing to protofibrils. We considered the known aggregation breakers KLVFF, GSGFK, and LPFFD. The Aβ protofibril was derived from PDB ID 2LMO extracting residues 16-35 from chains A-F to produce a U-shaped fibril. The choice of this protofibril was partially due to the requirements of ADCP, which only accepts peptides that are 20 amino acids or less in length. We assessed the impact of aggregation breakers on incoming Aβ peptide binding by running ADCP simulations with and without aggregation breaker peptide inhibitors present. Our approach therefore permits us to measure the change in energetics and molecular contacts of incoming Aβ peptides with the protofibril due to inhibitors. We demonstrate that ADCP is a useful tool to screen aggregation breakers and estimate their potential therapeutic benefit