PRESENTERS

Department of Computer Science & Engineering

ASDRP Autograder: A competitive coding web app

The ASDRP Autograder is a competitive coding web application designed to enhance programming education through an interactive, automated evaluation platform. The system allows users to submit code directly in-browser, where it is compiled, executed, and graded against dynamic test cases in real time. By integrating Firebase for storage, and supporting Google OAuth sign-in restricted to ASDRP members, the platform ensures secure participation. Features such as a live leaderboard, problem dropdowns, autosaving, and a CodeFlask-powered editor create an engaging user experience that simulates professional coding competitions. Beyond simply grading, the autograder fosters collaboration and friendly competition among students, making it an impactful tool for teaching, practice, and assessment in computer science education at ASDRP.

RESEARCHERS: Tithi Raval, Irvington High School, '26; Shaurya Jeevagan, Independence High School '26

ADVISOR: Liu Lab, Software Engineering

KEYWORDS:  Competitive programming | Web application | Automated Code Evaluation

Department of Computer Science & Engineering

Forecasting Time-Delayed Blood Dynamics in the Mackey-Glass System using a Time-Step Based Prediction Framework

Chaotic biological systems, such as blood concentration in leukemia, arrhythmias, and EEGs, present significant challenges in healthcare, as their unpredictability makes accurate forecasting and effective treatment difficult. Accurate prediction methods are therefore essential to enable hospitals and first responders to deliver personalized care. Traditional machine learning approaches rely on autoregressive forecasting, where past predictions are recursively fed into the model to generate future states. These methods compound error and obscure generalization ability for patients without historical data, limiting evaluation of long-term predictability.

In this study, we introduce a novel timestep-based formulation for chaotic sequence prediction. Our approach directly maps an initial condition and normalized timestep to the future state. Training on many trajectories with distinct initial conditions allows the model to learn underlying dynamics that generalize to unseen cases, eliminating dependence on prior observations. This also enables explicit measurement of prediction horizon, or the maximum timestep over which predictions remain reliable before divergence.

We evaluate this framework on the Mackey–Glass system, which was created to model blood dynamics in leukemia patients. Models were trained on multiple initial conditions and were tested on a held-out trajectory (x₀ = 1.2). Among the six tested architectures, reservoir computing achieves the longest reliable prediction horizon of 200 timesteps, while also maintaining an RMSE below 0.05 and the lowest overall error. The performance of these models was significantly better than traditional autoregressive forecasting. Our work provides a practical framework for selecting reliable forecasting models in chaotic biomedical systems, supporting clinical decision-making in hospitals and urgent care centers.

RESEARCHERS: Aarjav Jain, Milpitas High School '27; Hruday Nara, Leigh High School '26; Alden Raymond, Folsom High School '29; Karthik Natraj, Monte Vista High School '26

ADVISOR: Akl Lab Machine Learning for Condensed Matter Physics 

KEYWORDS: Mackey Glass | Chaos Prediction | Machine Learning | Clinical Application

Department of Computer Science & Engineering

Machine learning in electron microscopy image analysis: nanoparticles characterization 

Nanoparticles of a diameter less than 100 nm find applications in diverse fields such as semiconductors, nanotechnology research, quality control, biochemistry, and drug delivery. In this work, we developed and evaluated a machine learning model for nanoparticle identification and morphological identification. For the dataset, a 70/30 split was utilized for training and validation and incremental increases of size of the training set by multiples of ten were utilized to construct a learning curve. Training sets were tagged and labeled in order to allow model learning, with later performance assessment against validation sets. Across five runs, misidentified particle rates decreased from 202 to 24, representing a detection rate of approximately 90%. The model was found to be highly morphologically robust and correctly detected circles, squares, triangles, pentagons, and irregular ("blob/unknown") shapes. The findings herein show the potential of machine learning to accelerate the accuracy and effectiveness of nanoparticle characterization.

RESEARCHERS: Bryan Li, Woodberry Forest School '27; Simon Tchira, Dr. Michael M. Krop Senior High School '26; Siddharth Mahesh, Mt. Carmel High School '28; Andrew Kim, Corona Del Mar High School '26; Carter Tsao, Harvard-Westlake '27; Matthew Nadavallil, California High School '29

ADVISOR: Starostina Lab, Materials Science 

KEYWORDS: Machine Learning | Scanning Electron Microscope(SEM) | Nanoparticle | Spherical | Cubical

Department of Computer Science & Engineering

Evaluation of the Mechanical Properties of 3D-printed PLA in two printing orientations

Additive manufacturing is an alternative to traditional manufacturing, applicable to a range of fields from structural and civil engineering to dental and precision manufacturing. Recent studies show that the mechanical properties of 3D printed parts depend on printing parameters, color and brand of the feedstock and brand of the printer. This study thus aims to expand on and evaluate previous findings by comparing the mechanical properties of PLA samples with multiple printing orientations (vertical and horizontal), multiple sizes, and multiple printer types. Methods used include ASTM tensile testing and analysis of fractured surface morphology using the Keyence optical microscope. The results will be compared to the literature and discussed.

RESEARCHERS: Sindhuraa Selvam, Foothill High School '27; Sophie Cagdaser, Notre Dame San Jose '26; Annika Hedge, Leland High School '27; Hayagriv Sriram, Dougherty Valley High School '27; Sreeram Chadalavada, Bellarmine College Preparatory '27

ADVISOR: Starostina Lab, Materials Science 

KEYWORDS: Materials Science | Additive Manufacturing | Materials Characterization

Department of Computer Science & Engineering

Advancing Environmental Mapping and Forest Health Assessments with Drone Imaging

With the increasing concern for environmental conservation, there is a growing need for efficient methods of environmental mapping and forest health assessments. However, traditional methods employed by the U.S Forest Health Monitoring have faced controversy due to limited spatial resolution and integration of modern technologies. This research paper explores the application of machine learning algorithms in autonomous drones to conduct forest health assessments. Autonomous drones have the ability to collect timely, up-to-date data, which offers enhanced accuracy. This study focuses on training Deep Learning (DL) models to classify different environmental features based on aerial imagery captured by drones. To achieve accurate and efficient data collection, we will utilize Red-Green-Blue imaging and Convolutional Neural Networks (CNN) with the appropriate evaluation metrics, such as the Normalized Difference Vegetation Index (NDVI), Leaf Area Index (LAI), and foliage color, to create tree classes and identify forest health indicators. By integrating machine learning algorithms into forest health assessment, this study provides a more efficient, accurate, and up-to-date approach to monitor and evaluate the well-being of forests—supporting ongoing efforts towards environmental management and conservation.

RESEARCHERS: Aditya De, American High School '28; Mina Iqlas, Foothill High School '28

ADVISOR: McMahan Lab, Quantum Computing & Computer Science 

KEYWORDS: Autonomous Drones | Machine Learning | Forest Health Assessment | Environmental Mapping

Department of Computer Science & Engineering

A Physics-Guided Hybrid Neural Network for Predicting Diffusion Monte Carlo Energies in Ground State Stereoisomers

The energy of a molecule governs many fundamental biological and chemical phenomena—ranging from protein‑folding landscapes to therapeutic drug design. Traditionally, molecular energies have been obtained by solving the Schrödinger equation. However, due to its inherent inaccuracy for multi‑electron systems, approximations such as self‑consistent field (SCF) methods and Quantum Monte Carlo (QMC) approaches have been developed. QMC methods—and in particular Diffusion Monte Carlo (DMC)—achieve near‑exact ground‑state energies by stochastically projecting the N‑body wavefunction, making them a popular choice for high‑precision energy calculations. However, DMC scales cubically with the number of electrons, O(N³), rendering it prohibitive for larger molecules. To overcome this barrier, we develop a hybrid deep‑learning framework that combines a Graph Isomorphism Network (GIN) with a physics‑guided neural network to predict ground‑state DMC energies of stereoisomers within chemical accuracy (≤ 1 kcal/mol). The GIN extracts detailed geometric and spatial descriptors from atomic graphs, while the physics‑guided network fuses these embeddings with global atomic features (e.g., polarizability tensor, dipole moment) to output a DMC energy. We train our model on molecular structures and quantum‑mechanical properties from the QM7‑X dataset and generate reference DMC energies using QMCPACK’s DMC module to create the training targets. Finally, we benchmark the trained model against full DMC calculations to assess its predictive accuracy and computational efficiency. 

RESEARCHERS: Rishab Ghosh, Albany High School '26; Sreevatsa Prevela, Monta Vista High School '26; Krish Patel, American High School '26

ADVISOR: McMahan Lab, Quantum Computing & Computer Science 

KEYWORDS:  Molecular Energy | Diffusion Monte Carlo | Graph Isomorphism Network | Physics-Guided Neural Network | Polarizability Tensor | Dipole Moment

Department of Biological, Human & Life Sciences

Utilization of eDNA Analysis for Foothill Yellow-Legged Frog Population Studies.

The Foothill Yellow-Legged Frog(Rana Boylii) holds the status of endangered on both federal and state levels in the US, with California, its main habitat state, designating it as either threatened or endangered, depending on the "distinct population segments" that vary with location. Regardless of the title, it is undeniable that the population has plummeted since the late 1900s. In order to preserve such populations, one needs to locate a habitat and verify a presence, but traditional methods present several problems, with the find-and-capture method not only requiring permits, but also potentially disturbing the natural environment and breeding grounds, which would ironically be detrimental to the whole purpose of such an endeavor. In this project, the usage of water eDNA provides a non-invasive alternative to such a thing, as residue of mucus, skin, and feces that flow downstream suggest the presence of a target species, meaning the only thing needed is water samples and lab equipment. This work in progress aims to follow similar projects in order to confirm the presence of the Foothill Yellow-Legged frog in the Sunol Regional wilderness area or connected bodies of water, potentially being able to locate, study, and preserve the biodiversity of California.

RESEARCHERS: Iris Zhao, Emerald High School '27; Zahra Mottaghian, Los Altos High School '28; Aris Kao, Prospect High School '27

ADVISOR: Suresh Lab Conservation, Wildlife, and Environmental Science 

KEYWORDS: eDNA | Foothill Yellow-Legged Frog (Rana Boylii) | Population Studies | Conservation | Environmental Issues

Department of Biological, Human & Life Sciences

Genome Variation Analysis between Different American Populations to Improve Precision Medicine

Precision medicine, which focuses on how genetic variations affect human susceptibility to diseases and response to medicine, has a need for greater genetic diversity. Native Americans are an underrepresented group in genomic data and health care. This project studies data from 52 ancient Paleo-Eskimo samples and an ancient Alaskan Native infant sample from the European Nucleotide Archive. We have hypothesized that our analysis will yield notable clinical recommendations thereby reducing health inequity. We have conducted epidemiology research to identify major disease trends in our populations, performed variant calling on raw sequence data using Genome Analysis Toolkit 4, and identified significant pharmacogenetic recommendations through PharmCat and the PharmGKB database. Our analysis of the clinical recommendations and clinically-relevant variants we have found shows the need for precision medicine in our populations to provide more effective and equitable healthcare. 

RESEARCHERS: Laasya Vavilapalli, Emerald High School '27; Kanika Rawat, Notre Dame High School '26; Richa Prasanna, Notre Dame High School '26; Nicki Yazdi, Stratford Prep School '27

ADVISOR: Cunha Lab, Bioinformatics and Cancer Biology

KEYWORDS: Precision Medicine | Genome Analysis | Variant Discovery | Pharmacogenomics | Health Equity

Department of Computer Science & Engineering

Hybrid Quantum-Classical Graph Generative Models To Target Alzheimer's Disease

Contemporary drug discovery and development processes require billions of dollars and lengthy amounts of time, which is why researchers are utilizing computational chemistry methods like machine learning to speed up molecular synthesis pathways. Alzheimer's disease presents a specific challenge, as treatment requires molecules that can penetrate the blood-brain barrier (BBB). There are currently no effective cures for Alzheimer's disease. Our group is currently working on developing a generative adversarial network to generate chemically stable, novel, and druglike molecules that are of the right size and fit the molecular descriptors to penetrate the BBB. Not only this, but it must also inhibit the protein amyloid beta, which accumulates in the brain and is one of the main causes of Alzheimer's. To accomplish this we have implemented many algorithms to verify the molecule’s efficacy like a size checker and toxicity verifier. Furthermore, we have built our own generative adversarial network (GAN) to generate these molecules based off of the QM9 small molecule database. Once trained, this model has the potential to accelerate the discovery of promising therapeutics for Alzheimer's disease by shortening the amount of time needed in the lab to find a molecule with optimized properties for blood-brain barrier penetration and amyloid beta inhibition.

RESEARCHERS: Leena Adwankar, Irvington High School '27; Nitya Pisolkar, Archbishop Mitty High School '27; Akhil Muthyala, Emerald High School '28; Meera Minocha, Saratoga High School '27

ADVISOR: McMahan Lab, Quantum Computing & Computer Science 

KEYWORDS:  Drug Discovery | Computational Chemistry | Machine Learning | Generative Adversarial Network | Alzheimer's Disease

Department of Computer Science & Engineering

Utilizing Autonomous Drones to Identify and Mitigate Fire Hazards in Close Proximity to Power Lines in Remote Areas

In recent years, forest fires in the United States have risen sharply, with many ignited by damaged or downed power lines. This project proposes an automated drone-based system designed to assess the risk of such incidents by monitoring vegetation near power lines. A standard drone, outfitted with essential components, captures video footage of the target areas. The video is then broken down into still frames, which are analyzed using a custom-trained image processing model to detect hazardous vegetation or structural issues. If a potential risk is identified, the system transmits the data to a centralized database for review by local authorities. This early warning system aims to support timely intervention and reduce wildfire occurrences. Future testing will be conducted in parks and controlled environments to validate the system’s effectiveness and reliability.

RESEARCHERS: Nirupama Balaji, American High School '28; Isha Ramakrishna, Foothill High School '27; Elizabeth Ashley, Milpitas High School '26

ADVISOR: McMahan Lab, Quantum Computing & Computer Science 

KEYWORDS:  Wildfire Prevention | Automated Drones | Image Processing | Power Line Monitoring

Department of Chemistry, Biochemistry & Physics  

Evaluation of machine learning models for the classification of optimal coupling agents in diverse Amide Coupling Reactions

Reaction optimization is a very time, resource, and labor intensive process, as the optimal reaction conditions depend highly on substrate identity, thus requiring the repetition of the condition screening process. Additionally, the multidimensionality of the data makes it suited for an approach involving machine learning, which can help optimize reactions. Herein, we report a platform for standardizing and filtering open source reaction data from ORD (Open Reaction Database) and using this machine-readable dataset to train thirteen machine learning models, including linear, tree-based, kernel method, instance based, neural network, and ensemble architectures in the yield prediction and classification of coupling agents in amide coupling reactions. While yield prediction remained a difficult task for our models due to the complexity of our reaction data, our models performed with great accuracy when classifying reactions to their ideal coupling agent category, including carbodiimide-based, uronium salt, and phosphonium salt. To further validate this approach, we deployed our classification models on isoxazole coupling reaction data generated in our lab, and it successfully categorized the reactions by coupling agent type. Our results demonstrate that kernel methods and ensemble-based architectures perform significantly better than other models such as linear or single tree based. Additionally, molecular environment features, captured by XYZ coordinates, three-dimensional features, and Morgan fingerprints around reactive functional groups, boosted model predictivity more than bulk material properties such as molecular weight, LogP, and SMILES.

RESEARCHERS: Sourodeep Deb, Mission San Jose High School '26; Aarav Anand, Lynbrook High School '27

ADVISOR: Njoo Lab Synthesis | Physical Organic Chemistry | Catalysis | Chemical Biology | Spectroscopy | Medicinal Chemistry 

KEYWORDS: Chemical Reaction Optimization | Machine Learning

Department of Computer Science & Engineering

Design Thinking and its Application to Healthcare Innovation: an AI-enhanced Framework for Drug Discovery

The drug discovery process of new pharmaceuticals is costly, time-consuming and lacks patient-centered innovation. Artificial intelligence (AI)  has accelerated elements of drug development such as compound identification and toxicity screenings. However, it lacks adaptability and often detaches from users’ needs. Design thinking, a human-centered approach emphasizing empathy and ideation, can bridge this gap by reframing drug development challenges with patient needs in mind. The aim of this study involves integrating AI with design thinking as it optimizes efficiency while ensuring innovative and user-focused pharmaceutical solutions. By analyzing results from incorporating design thinking principles into prompt engineering with three AI models: ChatGPT, Gemini, and DeepSeek, this study explores how combining AI’s speed and predictive modeling with design thinking’s iterative approach can enhance drug discovery. The responses demonstrated a deeper understanding of the user's needs through the "Empathize" step and became more dynamic through the “Ideation” step. This research proposes a hybrid framework leveraging AI and the design thinking process ensures drugs are developed efficiently while prioritizing patient-centered care, ultimately leading to more effective and diverse pharmaceutical innovation.

RESEARCHERS: Jennifer Li, Mission San Jose High School '26; Neha Nagpal, BASIS Independent Fremont '26; Divya Raghuraman, Irvington High School '27; Ava George, Centennial High School '27

ADVISOR: Jahanikia Lab, Life Sciences, Neuroimaging, Psychology & Bioinformatics

KEYWORDS: Artificial Intelligence | Design Thinking | Drug Discovery | Healthcare | Large Language Models | User-Centered Design

Department of Computer Science & Engineering

Using Quantum Machine Learning to Improve Detection of Spinal Injuries, Defects, and Illnesses

Incorporating image identification machine learning models in the medical field has prospects for improving diagnostic accuracy and efficiency, quality of individualized treatment, and research productivity.  However, the complexity of the subject leaves many areas that have yet to be thoroughly and specifically researched, such as the classification of magnetic resonance images of spinal cords and injury detection. This research targets osteoarthritis, the abnormal growth, damage, and gradual degeneration of connective tissues, and its pathologies include osteophytes, foraminal stenosis, other forms of vertebrae misalignment or collapse that are likewise detectable through magnetic resonance imaging.  By employing a quantum convolutional neural network and visual transformer, we optimize classical computing solutions by taking advantage of the increased complexity, compression, and precision of quantum computing.  

RESEARCHERS: Mahika Reddy, Dougherty Valley High School '26; Avni Saxena, Irvington High School '27

ADVISOR: McMahan Lab, Quantum Computing & Computer Science 

KEYWORDS:  Machine Learning | Quantum computing | Medical imaging

Department of Chemistry, Biochemistry & Physics  

Synthesis and Evaluation of Carmofur Analogs as Membrane Rupture-Inducing Agents

Carmofur, a hexyl urethane analog of 5-fluorouracil, has demonstrated remarkable potency as an antineoplastic agent for colon cancer, and a covalent inhibitor to the main protease (Mpro) of SARS-CoV-2. Previously, our lab has demonstrated the utility of benchtop 19F NMR as a platform for high throughput condition optimization for a more scalable synthesis of carmofur (Wang, et al. Canadian J. Chem. 2023), which has furthered enabled the synthesis, evaluation, and identification of a library of carmofur analogs to explore structural diversification through side chain motifs and single-atom substitutions for both anti-SARS-CoV-2 Mpro and anticancer properties (Gu, et al. Discover Chem. 2025). Further, through evaluating the membrane-rupturing activity of small molecules using imaging of fluorescently labeled giant unilamellar vesicles (GUVs), we identified two lipophilic urethane analogs of carmofur bearing dodecyl urethane and octadecyl urethane side chains that have remarkably potent membrane-rupturing capability in the nanomolar range. This proposed a potential mechanism for the in vitro activities of lipidated 5-fluorouracil analogs, prompting us to further investigate the optimal alkyl chain length for maximal GUV membrane disruption and evaluate their potency against cancer cells. To further explore this structure-activity relationship, we synthesized a systematic library of 19 analogs consisting of carbamates, urethanes, and amides composed of varying alkyl chain lengths (C5 to C18) including branched-alkyl substituents. Ongoing efforts are focused on evaluating the membrane-disrupting activity of carmofur structural analogs through a series of biological assays, providing a foundation for future anticancer applications of carmofur and its analogs.

RESEARCHERS: Chancie Chou, Lynbrook High School '26; Lavernie Chen, Santa Clara High School '28; Kathleen Yi-ting Hsu, St. Francis High School '28

ADVISOR: Njoo Lab Synthesis | Physical Organic Chemistry | Catalysis | Chemical Biology | Spectroscopy | Medicinal Chemistry 

KEYWORDS: Forthcoming

Department of Computer Science & Engineering

Molecular TDA: Building Models to Personalize Drugs

Molecular TDA is an intersection between computational chemistry, mathematics, and AI. The goal of our project is to further the drug industry, personalize drugs, and decrease the risk of side effects. In this project, we are building a model to evaluate a dataset of diseases and their related drugs. Topological Data Analysis (TDA), a technique to understand the underlying structure of a dataset, will be utilized to help build the model. We aim to find similarities between different drug expressions to help the drug development industry. By identifying similarities, drugs can be better developed and specialized.

RESEARCHERS: Divya Raghuraman, Irvington High School '27; Pransh Dalal, Emerald High School '27; Minjee Kim, Leigh High School '26

ADVISOR: Jahanikia Lab, Life Sciences, Neuroimaging, Psychology & Bioinformatics

KEYWORDS: Topological Data Analysis | Kepler Mapper | Persistant Homology | Personalized Medicine | Drug Compounds | Dimensionality Reduction

Department of Computer Science & Engineering

Measuring Air Pollution with the use of Unmanned Aerial Vehicles

Greenhouse gases such as CO₂, CH₄, and NOₓ are powerful drivers of climate change and loss of biodiversity despite monitoring networks now available being costly, sparse, and incapable of resolving fine‐scale complexities in emissions. Using small gas sensors on unmanned aerial vehicles (UAVs) with machine‐learning models is thwarted by difficulties in sensor calibration, processing data for flight operations, and adaptive navigation but allows rapid, high‐resolution mappings of pollutant concentrations. We present a UAV system for monitoring pollutants here in which ground‐truths are used to train machine‐learning models for converting raw sensor data to right gas concentrations, adaptive navigation for targeted hotspots in flight operations, and high‐resolution mappings of emissions with a scalable cost‑effective solution over labor‑based monitoring with recommendations for priority work towards mitigation.

RESEARCHERS: Anirudh Rao, American High School '28; Alyssa Kwon, Dougherty Valley High School ‘27; Rohit Pulle, Washington High School '28; Yash Shekhawat, Mission San Jose High School '28

ADVISOR: McMahan Lab, Quantum Computing & Computer Science 

KEYWORDS:  Pollution, Autonomous Drones, Artificial Intelligence, Emission Reduction, Pollution Monitoring

Department of Chemistry, Biochemistry & Physics  

Synthesis, biological evaluation, and structure-activity relationship of diversified C-4 analogs of podophyllotoxin as tubulin inhibitors

The rich diversity of lignan small molecules derived from podophyllotoxin, a non-covalent tubulin inhibitor isolated from the Podophyllum family, has led to the clinical development of several FDA-approved anticancer agents, including DNA topoisomerase inhibitors etoposide and teniposide. While these compounds share the same tetracyclic core, two subtle structural changes that differentiate podophyllotoxin from its DNA topoisomerase-binding analogs—the presence of 4’ methylation on the aromatic ring and stereospecific glycosylation at the C-4 hydroxyl—changes the fundamental mechanism of which these two bioactive compounds act. Given the immense pharmacological importance of these two features, we sought to establish a structure-activity relationship between modification at C-4 and the potency, specificity, and chemical properties of podophyllotoxin. We previously reported with esters that increasing C-4 bulk decreases potency against a selection of human colon cancer cells but insignificantly impacts cell-free assays. Here, we report the first synthesis of a systematic library of 22 C-4 podophyllotoxin analogs with carbonate and carbamate substitutions. The antiproliferative activity and efficacy of our analogs as tubulin inhibitors were evaluated through in-vitro cell viability experiments, cell-free tubulin polymerization assays, cell cycle analysis, immunofluorescence imaging, and computer docking models.

RESEARCHERS: Shreya Somani, Lynbrook High School '26; Stella Yang, Harker '26

ADVISOR: Njoo Lab Synthesis | Physical Organic Chemistry | Catalysis | Chemical Biology | Spectroscopy | Medicinal Chemistry 

KEYWORDS: Organic Synthesis | Natural Products | Tubulin Inhibitors | Anticancer Therapeutics | Medicinal Chemistry | Chemical Biology

Department of Computer Science & Engineering

Investigating The Effects of Gamification on the Dual-N-Back Game: A Pilot Study.

The dual-n-back task is a cognitive training platform used to improve working memory functioning in individuals. However, users in the past have claimed that the interface is too mundane, creating a worse experience for them, thus discouraging them from training their working memory. Our group took the classic dual-n-back task and added popular game features to increase engagement. This is a great first step forward in providing engaging cognitive training to adults at risk for MCI (mild cognitive impairments) and other neurodegenerative diseases such as Alzheimer's, where training working memory has been shown to be a preventative measure. 

RESEARCHERS: Neha Sharma, Mission San Jose High School '26; Diigant Srivastava, American High School '26; Shriya Nanjanagud, Notre Dame High School '27

ADVISOR: Jahanikia Lab, Life Sciences, Neuroimaging, Psychology & Bioinformatics

KEYWORDS: Dual-N-Back Game | Cognitive Training | Working Memory | Neuroplasticity | Executive Function | Cognitive Enhancement | Brain Training

Department of Computer Science & Engineering

Quantum Error Detection and Mitigation

Quantum computing has the potential to significantly improve computational tasks. Unfortunately, errors due to outside interactions cause the data to be noisy. Our group is focusing on finding a way to decrease these errors by reverting the data back to what it was before the noise. To simulate realistic errors, we used a bosonic bath model on the surface code. We use both Convolutional Neural Networks (CNNs) for preprocessing and feature extraction, and Graph Neural Networks (GNNs) for predicting and correcting errors in the surface codes. By training and evaluating the CNNs and the GNNs, we can increase performance in quantum computing. These changes will allow for major developments in the field and will be a significant contribution to making an accurate quantum structure.

RESEARCHERS: Arhaan Reddy, The Quarry Lane School '28; Satvik Dronavalli, Independence High School '26; Vihaan Krishnakumar, Archbishop Mitty High School '26

ADVISOR: McMahan Lab, Quantum Computing & Computer Science 

KEYWORDS: Quantum Computing

Department of Chemistry, Biochemistry & Physics  

Investigation Into the Electronics of Para and Meta-Substituted Phenolic Esters for Selective Acylations of Amines

Esterase-cleavable prodrugs can generally be used to overcome poor pharmacokinetic properties as they can remain stable in the gastrointestinal tract until bioconverted into their active form by enzymes. This has been applied to mycophenolic acid (MPA), an immunosuppressant commonly used to prevent organ transplant rejection, for the preparation of mycophenolate mofetil (MMF), an FDA-approved ester prodrug derivative that enhances the bioavailability of MPA by around 90%. However, literature-reported synthetic pathways of MMF require multi-step syntheses and the usage of early-stage protections and hazardous catalysts such as thionyl chloride. Moreover, as the largest drain of atoms in large-scale chemical processes is not the reagents but rather the solvents, multi-step syntheses that require intermediary chromatographic purifications are economically and environmentally unsustainable on the industrial scale. To resolve this, we screen conditions to synthesize mono-siloxy MPA and install the mofetil substituent through single-pot esterification, averting the need for intermediary purification. We optimized a telescoped monosilylation, Shiina esterification, and HCl-mediated desilylation to provide MMF in a chromatography-free operation directly from MPA. By reducing the amount of solvent and hazardous reagents required in synthesizing MMF, we aim for optimized synthetic feasibility, allowing for maximized efficiency on industrial scales, thus increasing the accessibility of MMF for organ transplant patients. We envision this will provide an expedient synthesis for analogs of MPA.

RESEARCHERS: Abhinav Chalasani, Mission San Jose High School ‘26; Navya Sathish, Cupertino High School ‘26

ADVISOR: Njoo Lab Synthesis | Physical Organic Chemistry | Catalysis | Chemical Biology | Spectroscopy | Medicinal Chemistry 

KEYWORDS: Mycophenolic Acid | Mycophenolate Mofetil | Esterification | Immunosuppressants | IMPDH Inhibition | One-Flask Synthesis | Shiina Esterification

Department of Chemistry, Biochemistry & Physics  

Benchtop NMR enabled gram-scale, chromatography-free process optimization for the preparation of the mofetil ester prodrug of mycophenolic acid

Esterase-cleavable prodrugs can generally be used to overcome poor pharmacokinetic properties as they can remain stable in the gastrointestinal tract until bioconverted into their active form by enzymes. This has been applied to mycophenolic acid (MPA), an immunosuppressant commonly used to prevent organ transplant rejection, for the preparation of mycophenolate mofetil (MMF), an FDA-approved ester prodrug derivative that enhances the bioavailability of MPA by around 90%. However, literature-reported synthetic pathways of MMF require multi-step syntheses and the usage of early-stage protections and hazardous catalysts such as thionyl chloride. Moreover, as the largest drain of atoms in large-scale chemical processes is not the reagents but rather the solvents, multi-step syntheses that require intermediary chromatographic purifications are economically and environmentally unsustainable on the industrial scale. To resolve this, we screen conditions to synthesize mono-siloxy MPA and install the mofetil substituent through single-pot esterification, averting the need for intermediary purification. We optimized a telescoped monosilylation, Shiina esterification, and HCl-mediated desilylation to provide MMF in a chromatography-free operation directly from MPA. By reducing the amount of solvent and hazardous reagents required in synthesizing MMF, we aim for optimized synthetic feasibility, allowing for maximized efficiency on industrial scales, thus increasing the accessibility of MMF for organ transplant patients. We envision this will provide an expedient synthesis for analogs of MPA.

RESEARCHERS: Sarah Chang, Menlo Atherton High School ‘26; Oviya Srinivasan, Valley Christian High School ‘26; Kylie Yang, Archbishop Mitty High School ‘26

ADVISOR: Njoo Lab Synthesis | Physical Organic Chemistry | Catalysis | Chemical Biology | Spectroscopy | Medicinal Chemistry 

KEYWORDS: Mycophenolic Acid | Mycophenolate Mofetil | Esterification | Immunosuppressants | IMPDH Inhibition | One-Flask Synthesis | Shiina Esterification

Department of Computer Science & Engineering

Comparative study on three machine learning models in novel autonomous drone-based detection of invasive plant brassica nigra

California spends around $82 million to manage invasive plants each year (1). We propose a solution to automate the detection of invasive plant species by creating a machine learning model capable of identifying the presence of Brassica nigra - an annual herb which increases wildfire risk and produces chemicals that prevent the germination of native plants (1) - from autonomous drone footage. By recording the GPS location of this plant, we can locate Brassica nigra for removal. The drone consists of a Holybro X500 V2 ARF drone frame flown with a Pixhawk controller and other components. We tested three different machine learning models for the detection of the invasive plant from our drone footage at different angles and distances. The three models were a Convolutional Neural Network (CNN), Stochastic Gradient Descent Classifier (SGDC), and eXtreme Gradient Boosting (XGBoost), the latter being a gradient boosted decision tree. The goal was to find the best model type for this application. We hypothesized that for the detection of invasive plant species from aerial autonomous drone images, a CNN model will outperform SGDC and XGBoost because of its ability to extract spatial features to find complex visual patterns (2). Additionally, we hypothesize that SGDC will perform better than XGBoost, as our data is linearly separable and SGDC has the ability to do limited feature extraction. Results analyzed by using the values of the heatmap of each model indicate that there is a statistically significant difference between the ability of the three models to find important features with the ANOVA test, achieving a p value of 9.2e-16 at an alpha level of 5%. The drone records aerial footage of Brassica nigra using the GoPro’s, and a GPS location can be extracted. The methods included building and testing the drone, compiling an image dataset of about 200,000 augmented photos of both invasive and native plants from drone footage and the web, and training and testing the models. We can conclude that CNNs are the most suitable model for detecting invasive plants from drone footage due to its superior feature extraction abilities and that autonomous drones are an effective tool for collecting aerial footage to find invasive plant species over large areas in Northern California.

RESEARCHERS: Chloe Ho, Basis Independent Silicon Valley '26; Sahiti Pantangi, Washington High School '28

ADVISOR: McMahan Lab, Quantum Computing & Computer Science 

KEYWORDS: Autonomous Drone | Machine Learning | Brassica nigra

Department of Computer Science & Engineering

PsyColor Therapy: Investigating the Dynamic between Demographics, Personality, and Preferred Color

Color Therapy is the study of possible correlations between colors and their effects on human behavior. It explores the effect of different demographic factors, such as culture, age, and socioeconomic status on human reaction to different colors. Previous studies have explored the applications of color psychology in the realm of diseases and other physical ailments (stroke, fatigue), as well as the effect it may have on human emotion. This has also led to many other uses in the field of marketing. Additionally, since the colors we gravitate toward, whether in what we wear, buy, or surround ourselves with, often shift over time, it suggests that our preferences and personalities are dynamic and evolving. In this study, we aim to further investigate the effects of color psychology on physical and mental health, along with the impact of demographics on personality and favorite color.

RESEARCHERS: Joyce Yoo, The King's Academy '26; Tanu Goyal, Mountain House High School '26; Ajin Lee, Amador Valley High School '28

ADVISOR: Jahanikia Lab, Life Sciences, Neuroimaging, Psychology & Bioinformatics

KEYWORDS: Color Psychology | Demographics | Personality Tests | Color Preferences | Behavior

Department of Chemistry, Biochemistry & Physics  

Bridging the broad spectrum of chemistry: Amino triester lipids as biodegradable surfactants for drug delivery and precise control of quantum dot formation 

The delivery of anionic cargoes, including mRNA, small molecules, aptamers, and other oligonucleotide-based therapeutics, most fundamentally requires the enablement of a mono- or polyanionic payload to be delivered across a lipophilic membrane bilayer. Delivery systems typically rely on bifunctional cationic materials composed of a protonatable amine headgroup that electrostatically complexes with the desired anionic cargo, along with a lipid segment that permeates the hydrophobic lipid bilayer membrane. These materials may complex with anionic targets to form nanoparticles, liposomes, and other macromolecular structures, and these formulations have been previously described to be highly efficacious in mRNA vaccination gene delivery, siRNA delivery, and small molecular drug delivery. Given this, we synthesized–under nonhazardous, mild conditions– four non-toxic triester “tripod” lipids. Preliminary investigations demonstrated that these “tripod” lipids do not induce cell lysis, indicating potential for use in larger drug delivery systems. We then complexed the “tripod” lipids with calcium phosphate (CaP) in an attempt to create lipid-coated nanoparticles capable of localized drug delivery. Through a series of cell proliferation and fluorescence imaging assays conducted on lung and colorectal cancer cell lines, we evaluated the ability of our “tripod” lipids and our lipid-CaP conjugates in delivering clinically prevalent cancer therapeutic cargoes across the cell membrane. Lastly, we extended our study into the domain of inorganic chemistry by investigating the usage of our “tripod” lipids as surface ligands in the formation of quantum dots capable of fluorescent tagging in biological assays.

RESEARCHERS: Chancie Chou, Lynbrook High School '26

ADVISOR: Njoo Lab Synthesis | Physical Organic Chemistry | Catalysis | Chemical Biology | Spectroscopy | Medicinal Chemistry 

KEYWORDS: Organic Synthesis | Drug Delivery

Department of Biological & Life Sciences

Ensemble Learning Algorithms to Predict Scaffold/Matrix Attachment Regions

Scaffold/matrix attachment regions (S/MARs) are genomic elements that anchor DNA to the nuclear matrix, facilitating the organization of chromatin into structural and functional domains. These regions play a key role in gene regulation and DNA replication and mutations in S/MARs have been implicated in metastatic cancers. While previous studies have compiled a list of DNA motifs and sequences associated with S/MARs, there is yet to be a comprehensive, definitive database of S/MARs in the human genome. An accurate identification of S/MAR sequences could be a valuable resource for scientists studying cancer and other diseases. A robust and reliable method for S/MAR detection would also enhance our understanding of their regulatory functions and potential therapeutic implications. To this end, we implemented and optimized multiple machine learning algorithms for S/MAR detection; these algorithms included Random Forest, K-Nearest Neighbors (KNN), XGBoost, and Neural Network, all trained on 435 experimentally determined HeLa cell S/MARs from the ENCODE project dataset. As a negative control, inter-S/MAR sequences were utilized, and the models were trained on biologically relevant features such as AT-richness to improve classification accuracy. The Random Forest model achieved an 87.5% accuracy, the Neural Network yielded a 79.4% accuracy after 10 epochs, the XGBoost model resulted in an 83.2% accuracy, and the KNN achieved a 71.5% accuracy. These models are now being applied to identify novel S/MAR loci within the human genome. Further work is needed to fine-tune the hyperparameters of the models. A finalized R package will increase the accessibility of S/MARs-related research, facilitate studies on transcriptional regulation, and significantly advance research on S/MARs-related diseases.

RESEARCHERS: Sathvega Somasundaram, Evergreen Valley High School '26; Adithi Aia, Portola High School '26; Shreya Krishnakumar, Emerald High School '27

ADVISOR: Cunha Lab, Bioinformatics

KEYWORDS: Deep Learning | Bioinformatics | Computer Vision

Department of Chemistry, Biochemistry & Physics  

Blue fluorescent siloxytecans exhibit potent anticancer activity and enable direct real time quantification of intracellular uptake

Notwithstanding the natural abundance of silicon on Earth, silicon-containing compounds comprise of relatively few pharmaceutical drugs, though several have demonstrated greater bioavailability and lipophilicity in various natural products. Here, we apply this strategy in the preparation and biological evaluation of synthetic camptothecin analogs involving a C-10 silyl ether on SN-38, which not only blocks a site of metabolism but also imbues bright blue-fluorescent properties in such compounds. These siloxytecans exhibit comparable dose- and time-dependent antiproliferative activity in a broad panel of cancer cells. Uniquely, we demonstrate that the enhanced fluorescence of these compounds enables real-time, quantitative visualization of the dynamics and selectivity of intracellular uptake through fluorescence microscopy without the need for extensive sample preparation or installation of auxiliary fluorophores. We further demonstrate that the kinetics of cellular uptake observed by fluorescence microscopy are consistent with time-course washout experiments with subtle differences in anti-cancer potency in several cell lines. Further cell cycle analysis by flow cytometry and cell free topoisomerase inhibition studies suggests that these siloxytecans retain the topoisomerase inhibiting properties of camptothecin and other related topoisomerase I inhibitors. Collectively, these studies highlight the utility of quantitative fluorescence microscopy in investigating mechanisms of biological transport and anticancer activity of such siloxytecans.

RESEARCHERS: Sanika Vaidya, Lynbrook High School '26; Ashley Mo, The Harker School '26; Jessica Parvin, Leigh High School '26; Sripathy Sadagopan, American High School '26

ADVISOR: Njoo Lab Synthesis | Physical Organic Chemistry | Catalysis | Chemical Biology | Spectroscopy | Medicinal Chemistry 

KEYWORDS: Topoisomerase I Inhibitors | Siloxytecans | Silylated Natural Products | Fluorescence Microscopy

Department of Computer Science & Engineering

Simulating BB84 Protocol with Noise in Quantum Key Distribution Systems using IBM Qiskit

This study develops a simulation-based framework for Quantum Key Distribution (QKD) that incorporates a detailed noise model within the BB84 protocol. By systematically embedding real-world quantum noise factors, such as multi-photon emissions and channel disturbances, directly into the key exchange process, our approach enables a precise evaluation of security and efficiency. We introduce noise-aware security enhancements and mitigation strategies to counteract vulnerabilities, including potential eavesdropping scenarios. This methodology provides a rigorous, quantitative assessment of QKD performance under practical constraints, offering insights into improving the resilience of quantum cryptographic systems in real-world deployments. Ultimately, as quantum computing continues to break through existing cryptographic systems, our findings support the development of QKD systems as a viable, versatile, and scalable solution for securing the quantum internet and future communication networks.

RESEARCHERS: Aditya Das, American High School ‘26; Elaine Wang, Harker School ‘28; Samiha Das, Archbishop Mitty High School ‘28; Ashvik Gosh, American High School ‘26

ADVISOR: McMahan Lab, Quantum Computing & Computer Science 

KEYWORDS: Quantum Computing | BB84 protocol | Cryptography | Cybersecurity | Networks | Quantum Dots

Department of Computer Science & Engineering

Looking for Dyson Candidates

Dyson objects and other megastructures represent potential technosignatures of advanced extraterrestrial civilizations that could harness stellar energy through partial encasement of their host stars, producing detectable infrared excess emissions and unusual light fluctuations in astronomical surveys. The identification of authentic Dyson object candidates presents significant challenges due to the need to distinguish artificial signatures from natural astrophysical phenomena, but successful detection would provide unprecedented insight into the existence of Type II civilizations beyond Earth. Here, we present our analysis of 38,182 stellar observations from NASA's Exoplanet Archive using supervised learning techniques and outlier detection methods to identify 83 stars exhibiting anomalous infrared signatures across WISE photometric bands, as well as our development of conditional formatting protocols to establish key explanatory variables for future machine learning classification systems that could extend beyond the seven M-dwarf Dyson candidates previously identified by Project Hephaistos II.

RESEARCHERS: Prameet Guha, Neuqua Valley High School, '26; Sumedha Joshi, Dublin High School '26; Ditika Teckchandani, Mission San Jose High School ‘27

ADVISOR: Downing Lab, Data Science, Machine Learning, Astrophysics 

KEYWORDS: Astrophysics | Dyson Objects | Data Mining | Stellar Anomalies | Extraterrestrial Intelligence | Infrared Photometry | Technosignatures

Department of Chemistry, Biochemistry & Physics  

Discovery of A4P1W1, a fluorinated atropisomeric arylisoxazole acrylamide covalent inhibitor for the treatment of solid cancers

In the design of novel anti-cancer compounds, we integrate privileged scaffolds seen in the antibacterial space, notably an aryl-substituted 5-methyl-3-phenylisoxazole fragment, and a fully saturated nitrogen-containing heterocycle tethered to an acrylamide warhead, inspired by EGFR/BTK inhibitor, Ibrutinib, as well as other irreversible inhibitors such as Sotorasib, which targets KRAS G12C. This composed analog library utilizing the various fragments was tested on HCT-116, CT-26, Calu-1, SKOV-3, HT-29, MDA-MB-231, THP-1, HL-60, KOPN8, CD40L, 3T3, and HEK-293 cell lines to evaluate potency. The 2,6-chloro-fluoro-aryl isoxazole piperazine acrylate analog was observed to be the lead compound, exhibiting dose-dependent inhibition against various cell lines. To investigate the effect of alternate alkylating warheads, we further prepared four chloroacetamides bearing the different aryl fragments which showcased potent, broad, and non-selective inhibition of cell growth across the same panel of cancer cell lines.

RESEARCHERS: Lutecia Lam, BASIS Independent Silicon Valley '28; Sabrina Chau, Archbishop Mitty High School '28

ADVISOR: Njoo Lab Synthesis | Physical Organic Chemistry | Catalysis | Chemical Biology | Spectroscopy | Medicinal Chemistry 

KEYWORDS: Organic Synthesis, Medicinal Chemistry, Chemical Biology

Department of Computer Science & Engineering

Using Quantum Neural Networks (QNNs), Quantum Vision Transformers (QVT), and the Mathematical Morphological Reconstruction Algorithm (MMR) for Brain Tumor Detection

Brain tumors affect millions around the world, so detection is critical to helping doctors determine treatment. Currently, radiologists manually identify tumors through MRI (Magnetic Resonance Imaging) scans; however, this poses several limitations: it creates a heavy reliance on the experience of radiologists, has become increasingly costly and time-consuming, and is not as accessible to areas that lack the necessary resources and doctors. With the advancement of deep learning algorithms, a more accessible and efficient solution is possible. Given the existing research in classical Convolutional Neural Networks (CNNs) for tumor detection, Quantum Convolutional Neural Networks (QCNNs) and Quantum Vision Transformers (QVT) offer a promising approach to the problem. Mathematical Morphological Reconstruction (MMR), another image processing method, provides a relative metric for success in the QCNN, and is another classical alternative to CNNs. This research compares the accuracy and computational speed of the MMR, QCNN, QVT, and CNN algorithms to determine whether introducing a quantum aspect presents any noticeable advantage. To build these models, extensive datasets of MRI brain scans were collected. The MMR algorithm involved applying various techniques such as dilation, erosion, and skull stripping through OpenCV2's morphology functions. The QCNN algorithm utilizes quantum power to encode the data into a parametrized quantum circuit and apply convolutional and pooling layers. In terms of future steps, QVTs will be implemented with QCNNs for higher spatial understanding. So far, our results indicate that the MMR algorithm achieved up to 92% accuracy. These results will be compared with the accuracy of the QCNN, QVT, and CNN algorithms.

RESEARCHERS: David Chin, California High School ‘26; Mina Iqlas, Foothill High School ‘28; Zhongshi Wang, Valley Christian High School ‘26; Aditya De, American High School ‘28

ADVISOR: McMahan Lab, Quantum Computing & Computer Science 

KEYWORDS: Deep Learning | Quantum Computing | Computer Vision

Department of Chemistry, Biochemistry & Physics  

Reactivity-informed Pharmacophore Editing and Biological Evaluation of Andrographolide and Synthesis of A-ring Analogs: Closing the Loop on the Oxetane

Natural products and their analogs have long served as inspiration for the exploration and development of small molecules with therapeutic significance. One such compound is andrographolide, a labdane diterpenoid extracted from the plant Andrographis paniculata, which has been extensively studied as an anticancer therapeutic. It is known to function putatively through covalent inhibition of NF-kB, a transcription factor at the crossroad of a myriad of cell signaling pathways that modulate tumor survival and metastasis. Functionalization of the C-19 hydroxyl might alter the primary mode of action from inhibition of NF-kB to the modulation of the Wnt/𝜷-catenin signaling pathway. To interrogate the structure-activity relationship of this position, we synthesized a library of andrographolide analogs by protecting the C-19 hydroxyl with large, hydrophobic silyl and trityl ethers. Inspired by the observed biological trends amongst the library, we sought to further interrogate more complex A-ring oxy-functionalization. Among several targets for A-ring functionalization, we were intrigued by early isolation of an A-ring oxetane analog that is a biosynthetic byproduct isolated from A. paniculata (Jantan et al., Phytochemistry 1994), whose synthetic preparation and biological properties are not well described. After investigation of several synthetic route candidates, we identified an efficient route to access the A-ring oxetane with three chromatographic purifications. En route, we describe mechanistic insight into A-ring reactivity of andrographolide and its analogs. With the oxetane in hand, we then shifted our attention to further diversification of the A ring diol system, synthesizing analogs involving deoxygenation at various positions, and exploration of various oxidation patterns as viable intermediates to probe the reactivity of andrographolide.

RESEARCHERS: Rushika, Irvington High School ‘26; Galen Liu, San Mateo High School ‘26; Andrew Chyu, Dublin High School ‘26

ADVISOR: Njoo Lab Synthesis | Physical Organic Chemistry | Catalysis | Chemical Biology | Spectroscopy | Medicinal Chemistry 

KEYWORDS: Andrographolide | Natural Product | Oxetane