Click here for the full 2020-2021 journal.
Congratulations to our 2020-2021 CJSJ Authors!
Justin shen & davesh valagolam- PULMXNET: A NOVEL DEEP LEARNING ARCHITECTURE FOR THE DIAGNOSIS OF COVID-19 ALONGSIDE OTHER RESPIRATORY ILLNESSES (BACTERIA AND VIRAL PNEUMONIA, TUBERCULOSIS) FROM CHEST X-RAYS
The novel coronavirus (COVID-19) has brought tremendous international detriment as it has forced the world into a global pandemic and shutdown. COVID-19 and other respiratory illnesses directly impact the pulmonary tract of the human body, leading to noticeable structural differences which can be identified via a chest x-ray (CXR). This study highlights the implementation of a novel convolutional neural network (CNN) in the diagnosis of 5-classes of CXRs (normal, COVID- 19, tuberculosis, bacterial and viral pneumonia). This model was trained over 17 epochs and had 3,403,061 parameters. The overall 5-class accuracy for PulmXNet was 86.24%. The individual accuracy values for normal, COVID-19, tuberculosis, and bacterial and viral pneumonia were 85.59%, 85.45%, 84.16%, 89.33%, and 85.52%, respectively. This model was the first 5-Class CNN model to classify for normal, COVID-19, tuberculosis, bacterial and viral pneumonia CXRs and was able to maintain a relatively high accuracy and low parameters, demonstrating computational efficiency. Future investigations should look to further optimize this model with Bayesian or random search optimizations.
About the author: Justin Shen and Davesh Valagolam are seniors at Jericho Senior High School participating in Jericho’s renowned high school research program. Their research in machine learning has earned them a publication in an academic peer-reviewed journal. Justin’s recent innovations during the pandemic have earned him an invitation to guest lecture at NYU’s School of Professional Studies and recognition from news outlets like Newsday and News12 Long Island. Justin has also been invited to the United States Earth Science Olympiad’s national training program. For his research, Davesh has earned first place at the Long Island Science and Engineering Fair (LISEF), qualification for the International Science and Engineering Fair (ISEF), and first on Long Island for Computer Science in Junior Science and Humanities Symposia (JSHS). He has also showcased his research at Maker Faire. In exploring his enthusiasm for engineering, he has received numerous accolades, including qualification for the American Invitational Mathematics Examination (AIME) and the VEX Robotics World Championship.
Barbara Timmerman- ELUCIDATING THE ROLES OF SUBTYPES OF B CELLS AND FIBROBLASTS IN THE IMMUNE RESPONSE OF RHEUMATOID ARTHRITIS JOINT SYNOVIAL TISSUE
Rheumatoid Arthritis (RA) is an autoimmune disorder characterized by inflammation, while Osteoarthritis (OA) is developed through wear and tear on the body. B cells and Fibroblasts are activated by inflammation, however, the roles of subtypes of these cells in the progression of RA are unclear. Through single cell RNA sequencing on a dataset with RA and OA, the B4 subtype of B cells and the F2 subtype of fibroblasts were identified. The Immunoglobulin Heavy Constant Gamma 3 (IGHG3) gene, which codes for an immunoglobulin protein, was upregulated, suggesting its potential as a marker of RA. Through gene set enrichment analysis, upregulated pathways of the subtypes were identified. Immunoglobulin related pathways were upregulated in the F2 subtype suggesting that F2 cells are a crucial part of the immune response. Interferon related pathways were enriched in B4 cells suggesting that B4 cells regulate immunoglobulin class switching to Immunoglobulin G (IgG) molecules. The functions of the F2 and B4 subtypes are unique to RA and provide new information that is necessary to further elucidate the roles of B cells and Fibroblasts in RA.
About the author: Barbara Timmerman is a senior at Manhasset High School. She has found a passion for science research in biology and bioinformatics and enjoys participating in activities outside of school, such as volunteering at the local hospital and creating new STEM programs for Long Island Children’s Museum. In her free time, she enjoys dancing, singing, and participating in theatre productions.
Jennifer Lew- SAFEBUILD: THE RISK-BASED UTILITY POLE DESIGN SOFTWARE
On a yearly basis, California experiences fires, property damage, and prolonged outages due to the failure of electric utility infrastructure. In 2007, a Southern California Edison (SCE) pole broke, igniting the Malibu Fire. In 2011, in the San Gabriel Valley, 248 SCE poles broke, causing an outage to 440,000 customers for up to a week. In 2018, a component on a PG&E tower broke, igniting the Camp Fire, which destroyed 18,804 buildings and killed 86 civilians. These structural failures were due, in part, to the utility’s failure to calculate the probability that its poles and towers could withstand known local wind speeds without breaking. Existing pole design software, SPIDAcalc and Osmose O-Calc, have major flaws. They provide inaccurate wind modeling analysis due to their failure to account for material strength variability. Additionally, the software calculates a strength factor using a Reference Wind Load, which is of little practical value as it cannot be compared against known local wind speeds. The current research corrects these errors by introducing a program, SafeBuild, that, given a wind gust, provides the probability that a structure can withstand the wind gust without breaking.
About the author: Jennifer conducted her research alongside her mentor Mr. Derek Fong, PE, Senior Utilities Engineer Supervisor at the California Public Utilities Commission. She was inspired to do her research after she learned of the devastating consequences posed by falling utility poles. When a pole breaks, the falling conductors can ignite fires, shock pedestrians, and cause massive property damage. To help her community, Jennifer was determined to develop a transparent, risk-based methodology for the design of utility poles. She believes SafeBuild will make a huge difference in the world by preventing wildfires, saving lives, and reducing costs to ratepayers.
Benjamin Punzalan- THE EFFECT OF GINKGO BILOBA EXTRACT ON BETA-AMYLOID AGGREGATION IN C. ELEGANS
In 2019, the Alzheimer’s Association estimated that approxima tely 5.6 million people in the United States alone are affected by Alzheimer’s disease. Alzheimer’s is a neurodegenerative disorder associated with the accumulation of beta-amyloid proteins, resulting in inflammation that disrupts synaptic functioning. The purpose of this experiment was to determine the extent to which Ginkgo biloba extract could be used to remediate beta-amyloid-induced paralysis in C. elegans. Ginkgo biloba is an antioxidant containing ginkgolides, which reduce inflammation by regulating cytokines. Additionally, Ginkgo biloba contains flavonoids, which reduce oxidative stress by limiting the buildup of free radicals, another characteristic of Alzheimer’s. Transgenic strain CL2120 worms, which express beta- amyloid paralysis, were exposed to either 50μg/mL, 100μg/mL, 150μg/mL, or 0μg/mL (control) of Ginkgo biloba extract. After 48 hours, paralysis was determined by prodding individual worms with a platinum worm pick, with full body movement indicating the worm was not paralyzed, and no movement or only head movement indicating paralysis. Statistical analysis of the data using IBM SPSS v. 25 ANOVA followed by a Post Hoc Scheffe with p<0.05 showed that there was a statistically significant reduction in paralysis in worms treated with Ginkgo biloba extract when compared to the control. Overall, there is a correlation between the addition of Ginkgo biloba extract and a reduction in beta-amyloid- induced paralysis. This data may lend itself in the future to studies observing not only the use of Ginkgo biloba extract as a treatment for Alzheimer’s symptoms, but also as a method of Alzheimer’s prevention.
About the author: Ben is a high school sophomore currently attending Manhasset High School. He enjoys biology and chemistry, and has been extremely interested in studying Alzheimer’s disease treatment in model C. elegans. His goal is to apply his research to help people globally in effective and cost-efficient ways. Outside of research, Ben enjoys playing tennis and volleyball, and has a passion for studying music theory.
Broderick Nies, Nicole Tian, and Thendral Kamal- AN AUTOMATED CENSUS OF GLOBULAR CLUSTER SYSTEMS IN VIRGO CLUSTER DWARF GALAXIES
This study outlines a method to detect globular cluster (GC) systems within Virgo cluster dwarf galaxies with the aim of creating a robust new census of GCs. Compact objects are selected in each image with the Source Extractor software package, and color-selection is performed to pick out possible GC candidates — objects within each dwarf galaxy that match the color and brightness profile of known GCs. Additionally, artificial GC sources are added to the image, and the percentage recovered by Source Extractor is used as a means to quantify the degree of completeness of their detection algorithm. Combined with an analysis of the GC background and the known shape of the GC luminosity function, this study obtains a statistical estimate of the total number of GCs associated with several dwarf galaxies. Further analysis of the new clusters discovered using this method could provide insight into the hierarchical formation of galaxies and the tidal events that modify them within the Virgo Cluster.
About the author: Broderick Nies is a Senior at Ralston Valley High School in Arvada, Colorado. Broderick loves to read, compose music with synthesizers, and take pictures of the night sky. Nicole Tian is a junior at the Harker School in San Jose, California. She loves astrophotography and computer programming, and is an officer for her school astronomy club. In her free time, Nicole loves to draw and hike. Thendral Kamal is 18 years old, and is studying A-Levels in Chemistry, Physics and Maths at the Sharjah English School in United Arab Emirates. In 2018, her experiment to test buoyancy in microgravity conditions won the UAE Zero Gravity competition, and she successfully tested it onboard a parabolic flight. Thendral enjoys stargazing, scuba diving, and aspires to become the first Indian woman on Mars someday.
Lucy zha- INVESTIGATING THE THERAPEUTIC POTENTIAL OF CAPSAICIN AND CURCUMIN: A COMPARATIVE STUDY ON NEUROBLASTOMA AND HYPOTHALAMIC CELLS
Neuroblastoma is one of the most common malignant pediatric tumors. However, current cancer therapy usually involves surgical removal or chemotherapy, which are both damaging to normal cells. Phytochemicals are proposed as cancer treatment alternatives. Curcumin and capsaicin, two polyphenolic compounds from turmeric and pepper, possess anti-inflammatory and antioxidant properties, which suggest their neuroprotective benefits. This study investigates anti- cancer effects of curcumin and capsaicin on neuroblastoma cells in vitro while also probing potential effects of curcumin on hypothalamic neurogenesis in vivo. The experiments demonstrated that curcumin and capsaicin synergistically induced cell death, inhibited metastasis through decreasing MMP9 levels, and disrupted tumor growth of neuroblastoma cells. Both compounds also induced cell death of hypothalamic cells in vitro, and curcumin inhibited hypothalamic neurogenesis, possibly through the Notch signaling pathway. These results illustrate that while curcumin and capsaicin are effective at treating neuroblastoma cells, their toxicity on hypothalamic cells in vitro and potential impacts on neurogenesis in vivo must also be considered in conjunction with their anti-cancer effects in future research.
About the author: Lucy is currently a high school senior at the Wheatley School. With a deep passion in biology and chemistry, she aspires to find a treatment for neuroblastoma cells while attempting to reduce the drugs’ toxicity. Lucy was very honored to be selected into the United States Earth Science Olympiad program this summer, in which she discovered her passion for Earth Science. Alarmed by global warming and environmental degradation, she also has dedicated herself to find solutions for the global freshwater crisis that is accelerated by climate change. Through research, she has obtained insights on the interconnectedness of Earth’s ecosystems. Hence, she wishes to combine her knowledge in biology and environmental science to discover both clinical treatment for medical issues triggered by environmental contaminants and filter out the contaminants at the source.
Ericka lai- THE REMEDIAL EFFECT OF MUCUNA PRURIENS EXTRACT ON COPPER-INDUCED PARKINSONIAN BEHAVIORS IN DAPHNIA MAGNA
In 2018, the Parkinson’s Foundation estimated healthcare for Parkinson’s Disease (PD) (treatment, social security payments, lost income) to be a $52 billion industry. PD is a neurodegenerative disorder that induces difficulty in movement and reduced motor skills. The aim of this study was to determine the effects of exposure to Mucuna pruriens extract (MPE) on reduced locomotion as a result of copper-induced Parkinsonism in Daphnia magna. Further study of M. pruriens on D. magna served to validate its efficacy on multiple model organisms. Parkinsonian behavior was induced by exposing D. magna to copper sulfate (15.9 mg/L) for 96 hours. It was hypothesized that exposure to MPE would increase locomotion in D. magna that had previously exhibited Parkinsonian behaviors. D. magna were treated with MPE (20, 40 μg/mL) for 96 h. Their movement was then recorded microscopically for 1 min after each exposure. Movement was quantified with the use of wrmTrck, an ImageJ plugin. The distance each daphnid traveled and the frequency at which they spun in terms of bends was measured. Analysis of the data using IBM SPSS v. 25 ANOVA followed by a post hoc Scheffe (p < .05) showed that both groups treated with MPE significantly increased movement from the group treated with only copper sulfate. Frequency of spinning significantly decreased in groups exposed to MPE compared to daphnids only exposed to CuSO4. It was concluded that M. pruriens shows efficacy as a treatment for impaired behaviors found in Parkinson’s disease in D. magna.
About the author: Ericka Lai is a sophomore at Manhasset High School in Manhasset, NY. She is especially passionate about biology, chemistry and environmental science. Ericka is a USABO semifinalist, and her interest in biology has inspired her to research potential remediations for neurological diseases. In the future, she would like to research other areas of science and apply her findings to the real world. In school, Ericka participates in Mathletes and rowing. She plays the cello and likes to play multiplayer games in her free time.
Thomas Y. Chen- INTERPRETABLE CONVOLUTIONAL NEURAL NETWORKS FOR BUILDING DAMAGE ASSESSMENT IN SATELLITE IMAGERY
Natural disasters ravage the world's cities, valleys, and shores constantly. Having precise and efficient mechanisms for assessing infrastructure damage is essential to channel resources and minimize the loss of life. Using a dataset that includes labeled pre- and post- disaster satellite imagery, we train multiple convolutional neural networks to assess building damage on a per-building basis. We present a highly interpretable deep-learning methodology that seeks to explicitly convey the most useful information required to train an accurate classification model. Our findings include that ordinal-cross entropy loss is the most optimal loss function to train on and that including the type of disaster that caused the damage in combination with a pre- and post-disaster image best predicts the level of damage caused. Our research seeks to computationally aid in this ongoing humanitarian crisis.
About the author: Thomas is an alumnus of the research program of the Summer STEM Institute (a virtual program), where he met his mentor for this project, Ethan Weber. He has a passion for using AI and computer vision for humanitarian good. Thomas is proud of this project because it has the potential to save lives and property in the event of natural disasters.
Jierui Wang- PROTECTIVE EFFECT OF GALLIC ACID AGAINST AΒ NEURONAL TOXICITY AND NEUROINFLAMMATION FOR THE TREATMENT OF ALZHEIMER’S DISEASE
Alzheimer’s Disease (AD) is one of the leading causes of death and dementia in elderly people. Previous research indicates that AD may take its roots from Amyloid Beta (AB) aggregation and neuroinflammation. As 3,4,5-trihydroxy benzoic acid, gallic acid, demonstrates its potential to alleviate AD induced cytotoxicity and oxidative stress, this study researched the effect of gallic acid on pathological formation of AD. Gallic acid was tested against AB induced cytotoxicity, neurons apoptosis, and lipopolysaccharide (LPS) induced neuroinflammation, via a variety of biological assays, including cell viability assays, cell mitigation assays, and ELISA assays. According to the data and statistical analysis, gallic acid attennued the neuronal apoptosis by reversing the cytotoxic effects induced by AB. Moreover, gallic acid also reduced the production of AB by significantly down regulating the protein expression of Amyloid precursor protein. Gallic acid indicated its effects on neuroinflammation since it significantly inhibited LPS induced pro-inflammatory cytokines released from immune cells. All the results demonstrated that gallic acid has a positive impact on AD by inhibition of AB aggregation and AD induced cytotoxic effects, as well as anti-inflammatory effects. Overall, gallic acid gives new insight into the treatment of AD.
About the author: Jierui has been doing research related to Alzheimer's Disease since her sophomore year (2018). She and her mentor, Dr. Zhu Wei from SUNY Old Westbury, have done a lot of research regarding the treatment for Alzheimer's Disease together. Jierui is interested in this topic because throughout conversations with elderly people in her community, she noticed that neurodegenerative disease is destructive to their families and mental health. She picked gallic acid as her main research focus because, besides its great anti-oxidant effects, it is an affordable and accessible option for many families. Jierui won the prize of outstanding Alzheimer's research in the American Chinese Physician Association science fair in 2019, and will continue to research gallic acid's anti-AD effects, hoping to make her discoveries practical and beneficial to the medical world.
Tomosuke Yamaguchi and Alson chan- NANOHOLE ARRAYS IN ULTRA-LOW CONCENTRATION BIOSENSING
Millions of medical tests are performed annually in the United States. However , current medical testing infrastructure and methods are prone to delays and errors. Biosensing using plasmonic nanohole arrays (NHA) can meet the demand for accurate and accessible testing methods. NHAs are metallic nanoscale structures with voids that allow nano-sized particles to pass through while accumulating biomarkers on its surface. Using finite-difference time-domain (FDTD) simulations, a method to quantify electrodynamic interactions, researchers can observe the presence and concentration of accumulated biomarkers accurately and efficiently. This technique enables the detection of a variety of biomarkers, such as immunoglobulins and antigens, at ultralow concentrations. These detection results can be interpreted by medical professionals to diagnose diseases at early stages.
About the author: Alson Chan and Tomo Yamaguchi had the opportunity to research the applications of nanotechnologies on disease diagnoses last summer. Both students have a vested interest in social impact through technological innovations. Alson is a senior at Gunn High School in Palo Alto, CA, and enjoys learning about and developing passion projects, producing music, and reading philosophy in his free time. Tomo Yamaguchi is a junior at Santa Rosa High School in Santa Rosa, CA who enjoys calculus and competitive swimming.
Evelyn hur- DEVELOPMENT OF CACTUS OIL DISPERSANT SUITABLE FOR OIL-DEGRADING MARINE BACTERIA AND VARIOUS SEAWATER TEMPERATURES
Following an oil spill, dispersants are sprayed in large volumes onto the ocean surface and work by emulsifying oil particles. However, most chemical dispersants contain toxic properties and inhibit natural oil-degradation by bacteria. This study investigates the effects of a natural dispersant from Opuntia ficus-indica cactus mucilage on oil-degrading Marinobacter growth and oil emulsification by testing marine broth absorbance and qualitatively rating dispersion respectively. The Opuntia ficus-indica was selected due to its unique surface-active properties, allowing it to naturally enhance dispersion. The results indicate that the mucilage enhances oil-degrading Marinobacter sp, M. aestuarii, M. antarcticus, and M. maritimus populations. Microscopic inspection suggests that mucilage is an effective dispersant both alone and in combination with the Marinobacter. The mucilage was also equally effective at various temperatures and without certain standard pre- treatment from previous studies. Lastly, a cell viability test indicated that both acidic and neutralized pH cactus mucilage did not damage cell growth, and is thus safe to use in oceanic waters. In short, cactus mucilage is a natural nontoxic dispersant that promotes oil- degrading Marinobacter and works well at various oceanic conditions.
About the author: Evelyn Hur is a high school senior at Seoul International School. Through her passion for chemistry, biology, and physics, she has conducted numerous studies on environmental issues and bioremediation solutions. In the past, she has won numerous national and international awards including the Silver Prize for the Samsung Humantech Paper Award and awards for Genius Olympiad, Korea Stockholm Junior Water Prize, and Innovative Green Technologies & Movements Competition. In school, Evelyn is heavily involved in solo and orchestral flute performance and wildlife and climate change activism.
Pranav kirti, isaac chang, and madeline day- GENOME ANALYSIS OF A NOVEL PHOTOARSENOTROPH, RHODOBACTER SP. STR. ORIO
Photosynthetic arsenite oxidation (photoarsenotrophy) is an anoxygenic process where arsenite is used as an electron donor for growth. Through this process, bacteria oxidize arsenite (As(III)) to arsenate (As(V)), a less toxic substance. Extremophiles from hypersaline environments, which are difficult to grow and genetically manipulate, were previously used to determine that photoarsenotrophy is encoded by the arxB2AB1CD gene cluster. Here, we analyzed a novel freshwater bacterium, Rhodobacter sp. str. ORIO, for its potential to serve as a model organism in the study of photoarsenotrophy. We utilized NCBI BLAST to gather arx sequences, EBI Clustal Omega to construct phylogenetic trees, and InterPro, Pfam, Phobius, and SAPS to analyze ORIO’s arx gene products individually. Lastly, we searched for the arx pathway in over 35,000 metagenomes from JGI IMG/MER and mapped the presence of photoarsenotrophs worldwide. Our findings indicate that ORIO contains an arx gene cluster similar to those previously studied, and that arxA-like genes are ubiquitous in nature, with concentrations in the U.S. and Southeast Asia.In conclusion, these results suggest that ORIO can serve as a model organism for photoarsenotrophy – providing the first step toward bioremediation by detoxifying arsenic-contaminated water and improving water quality.
About the author: Pranav Kirti is a senior and aspiring scientist from Lynbrook High School. He fell in love with biology and science as a child and has immersed himself in research multiple times. However, his diverse interests span far beyond biology, including computer science, entrepreneurship, community service, and philosophy. Isaac Chang is a senior at Saratoga High School interested in microbiology and its real-world applications. He is a pianist for his school's advanced jazz ensemble and a member of his school's Science Bowl team. Madeline Day is a senior at Amador Valley High School with a passion for microbiology and public policy. She is the co-founder of her school’s Biology Club and is also her school’s public forum debate captain. In the future, she hopes to combine her two interests to make positive societal change.
yash narayan- DEEPWASTE: INSTANTANEOUS AND UBIQUITOUS WASTE CLASSIFICATION FOR COMBATING CLIMATE CHANGE
Accurate waste disposal, at the point of disposal, is crucial to fighting climate change. When materials that could be recycled or composted get diverted into landfills, they cause the emission of potent greenhouse gases. Current attempts to reduce erroneous waste disposal are expensive, inaccurate, and confusing. In this work, we propose DeepWaste, an easy-to-use mobile app, that utilizes highly optimized deep learning techniques to provide users instantaneous waste classification into trash, recycling, and compost. We experiment with several convolution neural network architectures to detect and classify waste items. Our best model, a deep learning residual neural network with 50 layers, achieves an average precision of 0.881 on the test set. We demonstrate the performance and efficiency of our app on a set of real-world images.
About the author: Yash Narayan is a junior at The Nueva School in San Mateo, CA. He is passionate about using computer science, machine learning, and robotics to solve humanity's most pressing problems. He is especially excited by the problems at the intersection between artificial intelligence and climate change. He is a six-time hackathon winner and his work has been featured in Vox, the SF Business Times, and the San Jose Mercury News.
yashika batra, je-won im, and nathan nguyen- GAMMA-RAY ANALYSIS OF THE MOST ENERGETIC BLAZARS TO PROBE THE COSMOS
Direct measurement of the Extragalactic Background Light (EBL) is difficult due to foreground emissions. An alternative method is to indirectly probe the EBL from its interaction with blazar gamma (γ) rays. The Fermi Large Area Telescope (Fermi-LAT) and H.E.S.S. collaborations proposed using a scaling factor ɑ to normalize EBL density based on previously existing models. However, initial normalizations analyzing 10 years of data from the Fermi-LAT Fourth Source Catalog Data Release 2 (4FGL- DR2) resulted in numerous outliers, whose values differed more than 3σ from an existing EBL model. We performed a new spectral analysis on 12 years of Fermi- LAT observations, focusing on outlier and bright sources. The changes of ɑ derived from our analysis resolve the issue for most of the outlier sources, while creating a new outlier from our “bright sources” sample. By estimating the factor ɑ for a large number of blazars, this study will contribute to the creation of a density map of the EBL.
About the author: Yashika Batra, Nathan Nguyen, and Jewon Im all participated in the University of Santa Cruz's Science Internship Program in the summer of 2020. Drawn together by their passion for astrophysics, they worked on a project that led them to a deeper understanding and appreciation for our Universe and scientific research. Yashika is attending Evergreen Valley High School (co 2021) and loves to sing, play the guitar, and bake. Nathan is attending Atholten High School (co 2022) and enjoys traveling, tennis, and playing the alto saxophone. Jewon attends Choate Rosemary Hall and likes to play the viola and do competitive programming. They continue to intern with Dr. Olivier Hervet, their mentor, hoping to contribute more to EBL research.
William tang, jennifer song, julia kim- OBSERVING THE RANDOM DISPERSION OF COMPARTMENTALIZED DROPLETS IN AN OIL-WATER EMULSION IN MICROGRAVITY
By injecting a droplet of oil into water, a contact surface forms between the two liquids, known as a liquid-fluid interface. The jamming of nanoparticle-ligands assemblies at this interface forms a permeable membrane. In microgravity, multiple droplets of a liquid can be suspended in another exterior liquid phase, forming a stable emulsion without coalescing of the interior droplets. The negligible force of gravity on the International Space Station (ISS) allows for the random distribution of these encapsulated droplets, whereas on Earth buoyant forces immediately affect the structures due to differences in density. On earth and the ISS, liquid-fluid interfaces were created by injecting a solution of POSS nanoparticles and silicon oil into aqueous polyacrylic acid, and their shapes were compared. In microgravity, the formation of fluid-liquid-fluid and fluid-liquid-gas structures further proved the stability of these systems in the absence of significant buoyant forces. The formation of stable, dispersed structures in liquid-fluid emulsions allows for a wide variety of applications including drug delivery, reagent encapsulation for on-demand reactive systems, and all-liquid batteries.
About the author: William, Jennifer, and Julia are researchers and students at Valley Christian High school (VCHS), in San Jose, California. As part of ISS STEM program at VCHS, they designed, assembled, tested and ran the experiment aboard the ISS under the mentorship of VCHS faculty. They enjoy playing Minecraft and doing competitive Physics in their free time.
eeshwar krishnan- A NOVEL WEIGHTED APPROACH TO PREDICT PROTEIN FOLD TYPE
Prediction of protein fold type is the first step in determining protein folding of amino acids. Predicting fold type is a difficult multi-class machine learning problem. Past work in predicting fold types has shown poor overall accuracy, although the methods worked well for determining some protein fold types. In this paper, we describe a novel framework to predict protein folding using a weighted approach, combining different machine learning approaches in a principled manner. This approach uses a weighted voting method to combine results from different machine learning methods to improve the accuracy of predicting fold type over individual methods. Results show an increase in the accuracy of protein fold measurements. Furthermore, the framework can be expanded to include new and emerging deep learning methods, and can serve to enable protein folding prediction.
About the author: Eeshwar is a junior at the University Scholars Program at the Pennsylvania Leadership Charter High School. His interests include computer programming, with a focus on machine learning, as well as robotics, and 3D printing. He is also a 2nd Dan Black Belt in Tae Kwon Do, the lead programmer and drive coach for his FIRST Tech Challenge robotics team, and serves as a mentor for his high school Vex Robotics team and middle school FIRST Lego League robotics team. This spring, he worked with his robotics team to 3D print and assemble PPEs (Personal Protective Equipment) which the team distributed to local healthcare workers and first responders. After graduation, he plans to study computer science.
rebecca zhu- PREDICTING MIGRAINES WITH MACHINE LEARNING AND FEATURE SELECTION
Often seen as severe headaches, migraines lower the quality of life for patients. To combat ineffective monitoring applications that can only provide information about previous attacks and not future occurrences, multiple types of machine learning models such as the Random Forest (RF), Logistic Regression (LR), and Support Vector Machine (SVM) models were trained to predict migraines occurrences. The data used was augmented by employing Synthetic Minority Oversampling Technique (SMOTE). By using feature selection methods, more discriminative features were selected to train the models. The RF model outperformed existing models with a classification accuracy of 0.9924 on the testing data.
About the author: Rebecca is a high school senior from Nashua, NH. As an avid researcher, she has conducted research at UMass Lowell's ECE Department on predicting migraines and network security. She was inspired to research about migraines after seeing many of her friends suffer from them. She hopes to positively impact her community through her research endeavors.
sanjna kedia and emily ma- THE REMEDIATION OF WASTEWATER USING A NOVEL MICROBIAL FUEL CELL WITH OPTIMIZED ELECTRICITY GENERATION AND AN ALGAE BIOREACTOR
The USA generates the maximum amount of wastewater per capita across the world, and according to the EPA, approximately $25 billion is annually spent on its treatment. In addition to the high costs, high residues of nitrogen and phosphorus are found in the remediated clean water. The purpose of this study was to create a scalable, novel microbial fuel cell (MFC)/algae bioreactor that would be more efficient in pollutant removal as well as energy consumption than current aeration technologies. The MFC consists of two chambers (cathode and anode) separated by a Nafion membrane. The control treatment, aeration, and MFC took 11 days, 3 days, and 0.9 days, respectively to remediate the water (90% dissolved oxygen increase). Additionally, the MFC was able to generate electricity at a sustainable voltage (0.62 V max). Anabaena biomass increase in the algae bioreactor effectively reduced nitrate levels. As shown in this study, MFC treatment holds promise for a more electrically efficient, time efficient and cost-efficient method for treating wastewater.
About the author: Sanjna Kedia and Emily Ma are current seniors at Manhasset High School in Manhasset, NY, where they teamed up to create this novel MFC/algae bioreactor device as part of their school Science Research program. Both students are passionate about environmental science and have pursued this field since 9th grade, and are intrigued about finding ways to protect the environment through renewable energy and water treatment. Sanjna is extremely involved in the STEM field, participating in science and math olympiads, and serves as vice president for her school’s green club. Emily is a highly competitive dancer and leads her varsity tennis team as a co-captain. Sanjna and Emily have received prestigious awards through this research including 2nd place at the International Spellman HV Clean Tech Competition, and have a patent pending on their device.
henrik torres, amith vasantha, helen chow, gepoliano chaves, PhD- SARS-COV-2 VARIANT ANALYSIS
The SARS-CoV-2 virus started the novel coronavirus pandemic. SARS-CoV-2 is an RNA virus that causes infection through the binding of the virion’s spike protein to a cell’s ACE2 receptor. The SARS-CoV-2 virion cleaves its way into the cell and deposits its RNA genome that hijacks the cell’s RNA replication system to produce more virions. During replication, genetic variance arises through single nucleotide polymorphisms (SNPs) that can enable a zoonotic jump or affect the transmissibility or lethality of a virus [3]. Our research focused on studying these SNPs from collected FASTQ and FASTA files of human, pangolin, and bat SARS-CoV-2 genomes on online databases such as the NCBI SRA Browser and GISAID and running files through variant call pipelines. Our results confirmed SNP frequencies at locations in the genome that matched those of Yin [1]. Genomic comparison of SARS-CoV-2 between the humans, bats, and pangolins showed a distinct mutation at location 23403 on the spike protein that was only present in humans, possibly a mutation that facilitated a zoonotic jump. Regional frequency analysis of collected samples showed regional clustering and similarities. The results from our research further existing knowledge of the SARS-CoV-2 virus and can be further expanded upon to create regional vaccines specifically tailored to mutations that affect certain protein mechanisms.
About the author: High School students Henrik Torres, Amith Vasantha, and Helen Chow spent the summer studying variants of SARS-CoV-2, the virus that caused the COVID-19 pandemic. Henrik is a junior at Choate Rosemary Hall in Wallingford, CT. He is passionate about biology and chemistry, reading, writing for publications, and learning about interdisciplinary applications of science and medicine. Amith Vasantha is a sophomore at BASIS Independent Silicon Valley in San Jose, CA. Amith is passionate about all the sciences, computer science, and the way science and biology interact. Helen Chow is a junior at Lowell High School in San Francisco, CA. Passionate about biology and chemistry, she is grateful that this research allowed her to analyze and understand mutations in genomes on a deeper level.
arya khokhar- DEEP BRAIN STIMULATION IN THE CORTICO-STRIATO-THALAMO-CORTICAL PATHWAY AND ITS EFFECT ON OBSESSIVE-COMPULSIVE DISORDER
Obsessive-compulsive disorder (OCD) is a neuropsychiatric disorder in which repetitive behaviors are done to relieve anxiety caused by repeated and intrusive thoughts. About 20% of OCD patients remain resistant to therapies and medications and are linked to suicidal behavior and lack of social functioning. Deep brain stimulation (DBS) has been considered as a last- resort solution for these patients. Recently, neuroimaging techniques have shown significant differences in the activity of the cortico-striato- thalamo-cortical (CSTC) pathway in OCD patients, supporting the first studies of DBS in the anterior limb of the internal capsule (ALIC), which is a part of this pathway. Since then, studies have expanded DBS into other locations of the CSTC pathway. With all these different regions being studied, many patterns have been found. However, as each location has a different degree of efficiency in each trial, the final goal should be to be able to determine which location will be most beneficial for patients. The purpose of this paper is to compare the studies and effects of DBS on OCD patients in varying parts of the CSTC pathway and discuss the goals and experimental setups of future studies to determine the best combination of stimulation parameters and DBS locations for patients.
About the author: Arya has always been fascinated by the human brain and artificial intelligence. Due to limited lab opportunities during this pandemic, she chose to write a review paper. Earlier this year, Arya developed a website for her community about COVID-19 which provided information on health risks and updates on vaccines and treatments. In the same mindset, she also founded an organization with the objective of empowering and motivating other students by sharing their stories. She has given talks about her own stories and experiences at nearby schools. When she is not doing school work, she not only enjoys expanding her knowledge about the brain but also loves to sketch with charcoal and play tennis.
Pranavi garlapati- SMALL MOLECULE INHIBITION OF ONCOGENIC KRAS AND DOWNSTREAM SIGNALING PATHWAYS IN PANCREATIC DUCTAL ADENOCARCINOMA
Common recurring mutations in certain cell signaling pathways continue to be found in the progression and metastasis of Pancreatic ductal Adenocarcinoma (PDAC), with Ki-ras2 Kirsten rat sarcoma viral oncogene (KRAS), a family of genes controlling control cell growth, being the most frequent one. Understanding the nuances of this mutation and its downstream effectors can provide a blueprint for effective treatment methods using small molecule inhibitors. This review will elucidate how activation of the mutated KRAS protein leads to the eventual activation of various intracellular pathways, leading to cell proliferation. In addition, new targeted therapeutic treatments in the form of small molecule inhibitors will also be discussed.
About the author: Pranavi Garlapati is a junior at the Texas Academy of Maths and Science and a research intern at the MD Anderson Cancer Center in Houston. She is most passionate about cancer biology and is currently researching remotely about the advancements in treatments of Pancreatic Cancer. Pranavi has always had a passion for research and has been in UT Austin's High School Research program as one of the two freshmen in the lab. Outside of science, Pranavi is also interested in public service through volunteering. She has received the Barbara James Service Award for over 100 hours of volunteering and runs a blog about advocating for the special needs community. She is the sole teen ambassador for the Dallas non-profit, Bryan's House and is also on the Youth Leadership Team for the North Texas Red Cross. Pranavi hopes to make a significant contribution to society through scientific research and service.