Abstract

This paper investigates the intersection of knot theory, topology, and molecular biology, with a focus on the formation and resolution of knots in DNA molecules. We introduce a mathematical model that captures
the topological transformations of DNA during replication and translation, integrating knot invariants such as the Jones polynomial and linking number to predict knot formation under various biological conditions. To support our theoretical framework, we present experimental results using gel electrophoresis and atomic force microscopy (AFM) to visualize knotted DNA structures. Our findings suggest that specific topological invariants, such as the Jones polynomial and linking number, may serve as predictive indicators of DNA damage and repair efficiency, potentially enabling topological diagnostics for genomic instability and cancer- relevant DNA damage.

Abstract

This study demonstrates that quantum strategies can resolve classical game-theory limitations, such as the cooperation breakdown in multi-player Prisoner’s Dilemma, and achieve new Nash equilibria unattainable through classical computation. Using QAOA and entangled Eisert-Wilkens-Lewenstein circuits, we simulate 2–4 player PD
and show that while QAOA accelerates payoff optimization, EWL circuits reveal new equilibri —including full cooperation at k = 2 & 4, where classical theory predicts universal defection. Our 4-player EWL grid sweep (19 million profiles, ~5-hour runtime) represents the largest such simulation to date. In the Schelling Point Game, entanglement boosts coordination from ~50% to 100%, even under strong biases. Although entangled sweeps are
slower than classical search for small k, they reveal equilibria that classical models cannot reach. Compared to prior work (e.g., Eisert et al.), this study expands the Nash search space, benchmarks runtime on near-term hardware, and proposes practical quantum coordination protocols for trade, logistics, and current events.

Abstract

In-flight medical emergencies (IMEs) occur in approximately 1 in every 6042 commercial
flights and represent a significant challenge in medicine in aerial settings. Current monitoring
solutions are expensive and is often unrealistic for individuals to purchase, and even for
hobbyist pilots with smaller aircrafts, it may not be sustainable for to buy the expensive
equipments by spending an obscene amount of money if they do not fly very frequently. This
study is centered around creating a cost-effective and portable in-flight health monitoring device capable of continuously measuring heart rhythms (EKG), heart rate, and blood oxygen saturation. The monitoring systems can be used as a baseline to predict any critical conditions and avoid any catastrophes. The prototype integrates a MAX30105 pulse oximeter, an AD8232
EKG sensor, and an Arduino Mega microcontroller, with C++ as our programming language. Performance was evaluated through comparative testing with professional medical equipment across a diverse sample population of different genders and ages. Statistical analysis, specifically through percent error, demonstrated an average accuracy rate of 98.06% for the heartbeat function and 98.23% for the blood oxygen function. These findings support the device’s potential as a practical and affordable tool for enhancing passenger safety and improving emergency response onboard an aircraft.

Abstract 

Anthropogenic methane (CH4) emissions cause roughly 700,000 premature deaths
annually; moreover, naturally-occurring CH4

is difficult to monitor and predict. Existing
methods lack the predictive accuracy and real-time response necessary for effective
mitigation, due to high predictive error. The goal of this study is to enhance CH4 forecasting and mitigation efforts through SPECO, a multiparametric, dual-stage system using a Quantum Long Short-term Memory (QLSTM) neural network for accurate CH4 hotspot prediction and sonophotoelectrochemical oxidation (SPECO) for CH4 oxidation. A physical oxidation chamber prototype was engineered using an ultrasonic transducer, yielding a 40%
decrease in CH4 as frequency increased. The novel introduction of ultrasonic cavitation was
shown to enhance CH4 oxidation by increasing reactivity through uniform propagation – an
emerging solution for future space energy. The multiparametric SPECO resulted in syngas
and H2 products used for recurrent energy feedstock, supported by computational validation. 

Abstract

We investigate the performance of quantum machine learning algorithms through

quantum support vector machines (QSVMs) using kernel methods applied to di=erent high-
dimensional biological data. Our study focuses
on evaluating performance as a function of training data size across both classical and
quantum SVMs, with feature space reduction performed via principal component analysis
(PCA). While classical SVMs tend to outperform QSVMs in extremely low-data
regimes, our experiments reveal statistically significant performance gains in favor of
QSVMs as training sizes increase, suggesting that quantum models may better capture
high-dimensional biological structure when more data is available. These results indicate
a promising advantage for quantum kernel methods in complex biomedical classification tasks as quantum hardware and sample sizes scale. It was also found that QSVMs were able to match or exceed classical SVM accuracy (>94% accuracy) whilst using 20–40% less training data, suggesting a potential advantage in data-limited biomedical settings. The findings also suggest that there is potential to lower data acquisition expenses by approximately $100,000 to $200,000 for datasets of standard scale.

Abstract 

As the need for higher amounts of processing power in Artificial Intelligence (AI) continues to
grow in complexity, traditional silicon-based CMOS logic gates face fundamental scaling
limitations at sub-5 nm nodes. This study examines molybdenum disulfide (MoS2), a promising two-dimensional (2D) semiconductor, as an alternative channel material in logic blocks essential
for energy-efficient AI microprocessors. Three CMOS inverter designs—pure silicon, pure
MoS2, and a hybrid configuration (MoS2 NMOS with silicon PMOS) were simulated using
Qucs-S and Ngspice under identical input conditions. Performance metrics including propagation delay, energy per switching cycle, and average power consumption were extracted and compared. Results demonstrate that while MoS2 transistors offer promising leakage advantages due to their excellent gate control and high on/off current ratios, their lower carrier mobility and higher contact resistance resulted in increased dynamic power consumption and slower switching
speeds compared to silicon. However, the potential for high transistor densities, as demonstrated by recent studies, indicates that MoS2 could play a significant role in future hybrid CMOS architectures. Continued research and development are necessary to overcome existing limitations and fully realize MoS2's benefits in energy-efficient AI microchips.