Colloquia Fall 2023 By Advisor

Akira Yamamoto

Characterization of FITC-Lysozyme by Multiple Modalities of HPLC

Noah Kaleekal • Sean Krachenfels • Vishnu Prashanth

Characterization of chemically modified proteins are important due to it's structure complexity. Advanced biochemical analysis is of increasing interest. The attachment of FITC to proteins allows for researchers to measure protein interactions, protein conformations, and protein localization when studying drug delivery systems and other cellular processes. Most commonly, Hydrophobic Interaction Chromatography (HIC)-HPLC has been used for modified-protein analysis. However, Reverse Phase (RP-HPLC) has been employed for characterization of small molecules. We use Lysozyme-FITC as a model molecule for study. Here we report the characterization of Lysozyme, a relatively small protein (14 kDa)-modified FITC using Reverse Phase (RP)-HPLC. Furthermore, we use Ion-exchange chromatography and ESI-MS to validate these findings. We find numerous peaks in RP-HPLC, with high resolution. Furthermore, the number of peaks suggest separation beyond variation in the Dye-Protein ratio. The great performance suggests potential utility of RP-HPLC for characterization of other chemically modified proteins. In the future, we look to characterize individual peaks using further protein analysis, such as LCMS.

Akira Yamamoto

A High-Performance Liquid Chromatography (HPLC) Method for Analyzing the Purity of y-Polyglutamic Acid using Ion-Pairing Reagents

Jazlyn Dias

Poly-γ-glutamic Acid (γ-PGA) is a naturally occurring biopolymer formed of repeating units of D-glutamic acid and L-glutamic acid. It is commonly found in Natto beans, a Japanese staple food fermented by the bacteria Bacillus Subtilis. γ-PGA has several applications in the medicinal, wastewater, and food industries due to its non-toxic all-natural form. This study focuses on a novel method to characterize the polymer γ-PGA extracted from Natto beans with respect to purity, chemical composition, molecular weight distribution, and other aspects. The polymer was extracted and purified through ethanol precipitation, and its purity was then measured through the utilization of the ion-pairing HPLC modality, which has not been reported yet for γ-PGA. In addition to the homemade γ-PGA, we also tested commercial grade γ-PGA in order to have a baseline to compare our results to. After testing several methods, we used Tetrabutylammonium chloride (TBAC) as our ion-pairing reagent in mobile phases consisting of water and methanol with a gradient of 0-100 (methanol and TBAC) over 20 minutes. We found that this method results in distinct peak formations, which can then be used to assess purity of the sample. The estimated molecular weight of the product was 103k-111k Da, as determined by size exclusion chromatography.

Bharat Poudyal

Shoot Proliferation Techniques: A Comprehensive Study on Ex-plant Regeneration of Some Fruit Species

Amanda Lee • Pranati Mannava • Ram Rishi Pagadala • Shambhavi Singh • Vihaan Chawla

In-vitro grown ex-plants are used for organogenesis and regeneration of genetically modified crops after gene transformation. This study assesses the shoot proliferation potential of nine fruit species - orange, mandarin orange, lime, guava, pear, plum, peach, grapes, and fig, which is a prerequisite for organogenesis and crop regeneration after gene transformation. Our objective is to identify species with high regeneration capacity, crucial for successful tissue culture techniques. Proliferation potential was investigated in MS media enriched with 6 Benzylaminopurine (BA) and Indole-3-butyric acid (IBA). Ex-plants, measuring 4-6 inches, underwent thorough preparation involving 75% leaf removal, followed by a two-hour tap water rinse. Sterilization procedures included a 10-minute treatment with 10% bleach and one minute with 70% ethyl alcohol. Ex-plants were subsequently washed with Deionized (DI) and sterile distilled water. Autoclaving was employed for 45 minutes at 0.12 MPa and 121°C to ensure the sterility of all equipment, while growth medium was subjected to a 20-minute sterilization process. Following initial challenges, the study successfully demonstrated the proliferation of peach, plum, and grapes in MS media supplemented with 1.0 mg/L BA and 0.1 mg/L IBA. These findings hold promise for future research in organogenesis within these fruit species. The investigation significantly contributes to the understanding of tissue culture techniques, offering insights into the proliferation capacities of various fruit crops. These results have profound implications for the advancement of propagation methods in horticulture research.

Clinton Cunha

Comparative Genomic Analysis of Colorectal Cancer Microbiome Bacteria to Discover Novel Relationships

Anjali Prabhu • Anish Jupudy • Shriya Viswanathan • Sumayyah Ismail • Harshita Keerthipati • Cheryl Cheung

Colorectal cancer (CRC) is uncontrolled tumor growth that starts in the rectum or colon (Park E. et al., 2022). Many factors affect the development of cancer, including daily habits, environments, and genetics. Our research focuses on analyzing the differences in pathways/enzymes between cancerous and non-cancerous associated bacteria in the gut microbiome outlined by a recent cancer microbiome review (Park E. et al., 2022). By utilizing the Bacterial and Viral Bioinformatics Resource Center (BV-BRC), we compiled our bacteria's genetic information into genome groups and used the comparative systems service to identify target pathways and construct phylogenetic trees. After focusing on genomes, we delved deeper into the enzymes. The programming language R was used to narrow down four specific enzymes from the set of genomes: two from the pathways only in non-cancerous bacteria and two in cancerous-associated bacteria. A Multiple Sequence Alignment (MSA) run at the genome level identified the range of lowest entropy among the genes in the four enzymes - one of which had the lowest range of 30-40. We are using NCBI Blast and other bioinformatics methods to characterize/validate the four enzymes in our respective target bacteria. Our end goal is to target/screen the unique pathways and enzymes (like the enzyme with EC number 5.4.3.2) of the cancer-associated bacteria and non-cancerous associated bacteria to decrease the metastasis of CRC tumors (Park E. et al., 2022). These genes, that help create the enzymes, can be manipulated in the wet lab as shown by the cited paper (Dong X. et al., 2022).

Clinton Cunha

In-silico characterization of potential serum response factor (SRF) inhibitors in colorectal cancer

Ojasvi Mudda • Aksithi Eswaran

Serum response factor is a transcription factor that is activated by growth factor stimulation and mitosis and plays a role in muscle tissue development. Additionally, in several types of cancer, SRF is overexpressed, causing the downregulation of E-cadherin which promotes the epithelial-mesenchymal transition. This results in metastasis and poor outcomes. We used chemical similarity algorithms and clustering techniques, like Tanimoto similarity and UMAP, to determine SRF inhibitor candidates based on limited existing inhibitors, comparing these candidates with the positive controls. Candidates from the CHEMBL database were docked to the target using Autodock Vina. Molecules with high binding affinities were tested for pharmacokinetic properties using SwissADME and drug toxicity with cell line data using Paccman. Furthermore, AutoGrow, an open-source program which uses a genetic evolutionary algorithm, will be used for de-novo drug design. Preliminary results reveal several inhibitors with better binding affinities than the positive controls. The BOILED-egg model generated through Swiss-ADME reveals that most of our candidates are unable to cross the blood brain barrier, but can be absorbed through the gastrointestinal tract, which shows they could potentially be administered to treat colorectal cancer. Additionally, one of our drug candidates developed from AutoGrow has the second highest binding affinity to SRF and the least number of drug interactions. We anticipate that these drugs will eventually be tested in-vitro on colorectal cancer cell models, along with expanding to other cancer like prostrate and gastric cancer by extending our drug toxicity models to different cancer cell lines.

Edward Njoo

Optimized Synthesis and Structure-Activity Relationship of Carmofur Analogs: Dual-Purpose Small-Molecule Inhibitors Against Human Acid Ceramidase and the SARS-CoV-2 Main Protease

Lexi Xu • Amber Lu • Tiffany Gu • Leyuan Zhang • Jade Lee • Anirudh Raman

During the COVID-19 pandemic, carmofur, a 5-fluorouracil derivative initially developed as an antineoplastic agent against colorectal cancer, was identified through drug repurposing as a potent covalent inhibitor of the SARS-CoV-2 main protease (Mpro), making it a promising therapeutic candidate against COVID-19. Previously, we reported the optimization of the synthesis of carmofur through using Benchtop 19F NMR to quantitatively track reaction rates in various reaction conditions. In this study, we developed an optimized protocol for producing six novel analogs of carmofur to explore the impact of structural modification on biological activity, bearing diversified alkyl, cycloalkyl, and aryl substituent chains. Additionally, in order to probe the effects of single-atom substitution at the electrophilic carbonyl on the carmofur side chain, we selected a heptyl amide analog and hexyl carbamate analog to synthesize wherein the nitrogen atom in carmofur is substituted with a methylene or an oxygen. Through monitoring reaction rates with 19F NMR, we were able to rapidly determine optimal synthetic conditions for each analog. To evaluate the efficacy of our compounds in inhibiting SARS-CoV-2 main protease activity, we performed a colorimetric assay to evaluate their inhibitory effects on the main protease. To probe the antiproliferative effects of our compounds, we performed MTT assays against colorectal cancer cells. All together, we demonstrate the utility of a 19F NMR enabled workflow in the preparative synthesis of novel 5-fluorouracil analogs, several of which exhibit selective potencies against the main protease of SARS-CoV-2 or as antiproliferative agents that are at times comparable or superior to carmofur.

Edward Njoo

Evaluation of Bio-inspired Ionizable Lipids for Lipid Nanoparticle mRNA Delivery.

Shreya Somani

Lipid nanoparticles (LNPs) have demonstrated exceptional promise as one of the most clinically advanced non-viral vehicles for messenger RNA (mRNA) delivery in vivo, as seen in recent developments of FDA-approved mRNA vaccines against SARS-CoV-2 infections. Typically composed of four main components–PEGylated lipids, phospholipids, cholesterol, and ionizable/cationic lipids, LNPs depend on branched ionizable lipids to play a major role in protecting the mRNA and improving cytosolic delivery. When in acidic pH, ionizable lipids are positively charged and spontaneously complex with polyanionic nucleic acids in the formation of LNPs; at physiological pH these ionizable lipids become charge-neutral, facilitating endosomal escape upon cell intake and protonation. However, most current clinically-available ionizable lipids require multi-step synthetic routes to manufacture. As a potential solution, we focused our efforts on designing a short, highly scalable, and efficient synthesis capable of producing biofriendly, non-toxic ionizable lipids inspired by nature. Here, we report a one-step synthesis producing a library of twenty-two lipids bearing unique naturally-derived acids containing linear and branched lipid tails, following an ionizable tertiary amine head group. Preliminary studies suggest that our ionizable lipids efficiently encapsulate FITC-labeled mRNA into nanoparticles and may be viable candidates for nucleic acid delivery. Further, we rationalize these results with in-silico modeling of the biophysical properties of these lipids.

Edward Njoo

Synthesis and Anticancer Activity of Close-In Analogs of Podophyllotoxin

Breanna Lu • Lexi Xu • Harriet Chen • Grace Yu • Kimberly Khow • Stella Yang • Arushi Dinker

While natural products themselves only make up four percent of FDA-approved small molecule therapeutics, analogs inspired by or derived from natural products account for over 40 percent. Among these analogs, slight structural changes to the biologically significant sites of podophyllotoxin, a clinically relevant natural product extracted from the plant Podophyllum, have been studied extensively for their anticancer and antiproliferative properties. Thus, we selected to synthesize seven novel close-in analogs at the C-ring hydroxyl of podophyllotoxin, which is a main site of previously reported metabolic degradation. At the C-4 hydroxyl, we explored the impact of steric hindrance on biological activity and mechanism of action through a systematic progression of ester analogs and the addition of a phenylthiol carbonate. Furthermore, we performed photolysis assays on photoreleasable o-nitrobenzyl photocages that selectively activate the C-4 hydroxyl and found that the addition of 4- and 5- methoxy substituents redshifts the wavelength for release and that the formation of a chromogenic byproduct stalls photorelease. We evaluate the activity of our analogs through cell viability assays, flow cytometry, tubulin polymerization experiments, and computer modeling of docking models. Our results show that large, aliphatic esters at the C-4 position are not well tolerated, leading to a precipitous loss of tubulin inhibitory and cancer cell antiproliferative activity. Additionally, installation of nitrobenzyl carbonates at C-4 attenuates potency and can be effectively released with the right wavelengths of blue LEDs, indicating that this may be a viable prodrugging strategy.

Edward Njoo

Fragment-based Discovery and Anti-cancer Activity of a Novel 5-methylisoxazole Piperazylacrylate Covalent Inhibitor for Broad Targeting of Key Oncogenic Markers

Polina Bortok

Covalent inhibitors have reemerged in the field of medicinal chemistry as previous concerns of off-target side effects were addressed through analysis of the specific protein binding pockets to determine the inhibitor's structure activity relationship (SAR). Acrylamide warheads have been identified to efficiently bind to cysteine residues and are prevalent in many FDA-approved covalent inhibitors such as Sotorasib against G12C mutant KRAS and Afatinib against Epidermal Growth Factor Receptor (EGFR). This study aims to analyze bioactive fragments in combination with acrylamide warheads and their SAR in improving covalent inhibition of oncoproteins. Heterocycles make up more than 85% of bioactive molecules, and the 5-methyl-isoxazole heterocycle fragment had been previously utilized in medicinal drugs such as the FDA approved Dicloxacillin used to treat bacterial infections. The shown bioactivity of this molecular fragment gives it potential to aid in covalent inhibition. Inspired by this, we developed a library of 12 isoxazole-based acrylamide inhibitors. Although the results of computer modeling suggested that our molecules would inhibit G12C mutant KRAS, these compounds were also found to be potent in in vitro antiproliferative activity in human colorectal cancer cells lacking the KRAS G12C mutation, suggesting that these molecules might act through an alternative pathway. Further inspection of computer models and docking simulations suggested that these compounds may covalently inhibit other cancer targets, such as EFGR. With these preliminary biological results in hand, we further investigate the primary pathway of action of these compounds through flow cytometry, transcriptomic analysis of key markers by RT-qPCR, and protein expression mapping.

Edward Njoo

Butyrylcholinesterase Inhibitory Activity of Rivastigmine Analogs

Stella Yang • Arshia Desarkar • Mirabelle Feng • Melanie Tsui • Mihira Gutti • Jennifer Luo • Aglaia Lan • Terry Wang

Neuroactive alkaloids isolated from naturally occurring sources have been used to identify small molecules that treat neurological disorders for centuries. Physostigmine, a neuroactive alkaloid that is active in the plant Physostigma venenosum, has been reported to be a potent cholinesterase inhibitor and is currently being used to treat neurodegenerative diseases such as schizophrenia and dementia. However, due to its low bioavailability, sixteen-minute half-life, and high toxicity, physostigmine has limited therapeutic potential. Our study focuses on rivastigmine, a synthetic compound that is structurally analogous to physostigmine. We synthesized multiple analogs of rivastigmine-bearing alkyl substituents of various lengths, and we previously reported that analogs with less steric bulk were more effective in inhibiting acetylcholinesterase (AChE). To further this study, we then evaluated this library of analogs in inhibition to butyrylcholinesterase, hypothesizing that this trend may be reversed given the greater steric bulk of BChE's native substrate. The structure-activity relationship (SAR) of these analogs possessing methyl, ethyl, and morpholino carbamates on both inhibitions of AChE and BChE will be presented as well as preliminary studies of comparing colorimetric versus mass spectrometry based assays in precise evaluation of cholinesterase activity.

Edward Njoo

Design, synthesis, and biological evaluation of C-4 ester analogs of podophyllotoxin

Breanna Lu • Grace Yu • Harriet Chen • Kimberly Khow

The rich diversity of lignan small molecules derived from podophyllotoxin, a non-covalent tubulin inhibitor isolated from the podophyllum family, has led to the discovery and clinical development of several anticancer agents including etoposide and teniposide, FDA approved DNA topoisomerase inhibitors. These two compounds, along with numerous other analogs of podophyllotoxin, exhibit unique biological activities and mechanisms of action due to modification at the C-4 hydroxyl. Given the immense pharmacological importance of this feature, we sought to establish a structure-activity relationship regarding modification at C-4 on the potency, specificity, and chemical properties of podophyllotoxin. We synthesized and evaluated a systematic library of close-in diversified esters at the C-4 position of podophyllotoxin to evaluate the effect of bulk on potency. We demonstrate the hydrolytic stability of our analogs through 24-hour esterase assays and evaluate their biological target and activity through cell viability assays, tubulin polymerization assays, and cell cycle analysis. Furthermore, we rationalize our results by analyzing the interactions between each ester and the binding site of tubulin through computer docking models. Altogether, our results show that increasing steric hindrance at C-4 leads to a loss in tubulin inhibition and antiproliferative activity against human cancer cells.

Edward Njoo

Reactivity-informed Pharmacophore Editing and Biological Evaluation of Andrographolide and its Analogs

Anushree Marimuthu • Tiffany Gu • Alice Tao • Carina Zhou • Yilin Fang • Emily Shu • Sanghyuk Ko • Srishti Venkatesan

Andrographolide, a natural product labdane diterpenoid extracted from the plant Andrographis paniculata, is known to have potent anti-cancer activity. The putative mode of action of andrographolide is the inhibition of Nf-kB, which subsequently leads to downregulation of a myriad of cell signaling pathways typically involved in cell cycle regulation. However, previous research has suggested that chemical modifications to the C19 hydroxyl may alter the biological target of the andrographolide analog to the Wnt/𝜷-catenin signaling pathway. With this pharmacophore in mind, we designed a library of targeted andrographolide C19 analogs with altered polarity and steric profiles to probe the effects of large, hydrophobic silyl and trityl ethers at C19 on metabolic stability and primary mechanism of action. After studying the potency of our analogs through MTT and cell migration assays, we assessed the analogs' downstream transcriptomic effects on key apoptosis-regulating pathways and their potential as a protein inhibitor in the Wnt/𝜷-catenin signaling pathways. Additionally, we installed a number of benzyl acetals to the a-ring of the andrographolide scaffold. These functionalities, being hydrolyzable under mildly acidic conditions, led us to hypothesize that they would function as hydrolytically labile prodrugs of andrographolide. In order to study the reaction dynamics of this process, we synthesized a library of 4-substituted benzaldehyde acetals and present a hammett linear free energy relationship of these prodrugs. We determined that the 4-chloro is the most hydrolytically stable.

Edward Njoo

Discovery, Process Optimized Synthesis, and Anti-Cancer Activity of 5-Phenylisoxazole Based Covalent Inhibitors Targeting G12C Mutant KRAS

Arshia Desarkar • Polina Bortok • Zoe Lin • Vera Lin • Natalie Brahan • Shreya Somani • Samuel Huang • Avni Sachdev

Oncogenic mutations in the GTPase protein KRAS are implicated in approximately 25% of human cancers— specifically, the most common mutation— G12C— is found in 13% of lung cancers. This single residue substitution causes irreversible binding to the GTP substrate, thereby forcing the protein into a permanent, activated state. While KRAS has been previously considered an undruggable chemotherapeutic target, the discovery of acrylate-based covalent inhibitors of G12C KRAS has led to the development of two FDA-approved chemotherapeutic agents: Sotorasib (AMG-510) and Adagrasib (MRTX849). Inspired by the pharmacophore model of these two compounds, we focused our efforts towards the development of a more efficient, economical, and diversifiable synthetic route. Through our efforts, we have developed a highly scalable three-step synthetic sequence, producing a library of twelve novel analogous isoxazole-based covalent inhibitors of G12C mutant K-Ras. En route, we optimized a previously-reported amide coupling that had previously suffered from long reaction times and requirements for super stoichiometric benzotriazole catalysts. Finally, potency of our compounds were evaluated in vitro against Calu-1 and MCF7 human cancer cell lines. Additionally, we rationalize our observed structure activity relationship (SAR) using computer docking models. Ultimately, this newfound SAR sheds new light on covalent targeting of G12C mutant K-Ras and related pathways connected to the activity of our lead analogs.

Edward Njoo

Synthesis and Anti-Inflammatory Activity of Prodrugs of Dexamethasone and Related Fluorinated Corticosteroid Analogs

Grace Yu • Jennifer Luo • Aglaia Lan • Terry Wang • Seoyeon Hong • Cindy Zou • Selina Xi • Divya Gottumukkala

Since the initial success of fluorinated corticosteroids in the 1950s, several, including dexamethasone, triamcinolone acetonide, and betamethasone have been investigated for their potent anti-inflammatory activities and improved pharmacokinetic profiles. Among these, dexamethasone received FDA approval in 1988 and has been prescribed to millions for the treatment of inflammation, arthritis, asthma, and multiple sclerosis. Prednisone, a similar corticosteroid, is delivered as its oxidized prodrug, prednisolone, to improve its metabolic properties with a shorter half-life. Inspired by this, we attempted a similar synthetic strategy on dexamethasone. We hypothesize the oxidation of its secondary alcohol at C-11 into a ketone might be an effective prodrugging strategy, wherein 11β-Hydroxysteroid dehydrogenase will dehydrogenate the ketone back into an alcohol. These compounds were evaluated for metabolic stability through HPLC and liver microsomal assays. Subsequently, we investigate methods of synthesizing a library of silyl-ether dexamethasone analogs and examine these compounds' gene expression activity relative to dexamethasone itself. While computer docking models indicate that silylation of the primary alcohol on the D-ring of dexamethasone leads to lower binding affinity, additional hydrophobic contacts with the corticosteroid receptor are observed. We subsequently synthesized two silyl ethers: a tert-butyl dimethyl and a tert-butyldiphenyl silyl analog. These silyl analogs are subsequently compared for activity in modulation of key inflammation-linked markers, including annexin induction. In a preliminary RT-qPCR transcriptome analysis panel, we determined that dexamethasone administration downregulates PTGS2 and TNIP3, two genes implicated in pro-inflammatory pathways, while upregulating REL. Parallel studies on our analogs are currently underway in our laboratory.

Edward Njoo

Optimizing the Synthesis of Novel Carmofur Analogs for In Vitro Evaluation as Dual-Purpose Inhibitors Of Human Acid Ceramidase and the SARS-CoV-2 Main Protease

Amber Lu • Lexi Xu

During the COVID-19 pandemic, drug repurposing efforts resulted in the discovery of carmofur, a 5-fluorouracil derivative originally developed as an antineoplastic agent, as a potential covalent inhibitor of the SARS-CoV-2 main protease (Mpro). Previously, we improved the lengthy and low-yielding synthesis of carmofur using benchtop 19F NMR. Here, we aim to enhance the selective potency of carmofur with analogs optimized to inhibit the Mpro active site. With benchtop 19F NMR spectroscopy, we were able to track reaction rates under different conditions by quantitatively monitoring compound concentrations over time. This enabled the rapid determination of optimal reaction conditions for carmofur analogs possessing different alkyl, cycloalkyl, and aryl substituent chains as well as for single-atom substituted heptyl amide and hexyl carbamate compounds. Our six novel analogs aim to probe the impact of diversified hydrophobicity, steric hindrance, and electrophilicity on biological efficacy. In-vitro studies on Mpro inhibition established that analogs bearing aryl substituents like the benzyl isocyanate analog and elongated alkyl chains like the dodecyl isocyanate analog had comparable potency to carmofur with EC50s of approximately 30 µM, whereas modifications to electrophilicity of the side chain did not result in significant inhibition. We also probed the antiproliferative effects of our analogs against colorectal cancer cells, where we found comparable or superior activity by the heptyl amide and hexyl carbamate analogs with IC50s of 14.0 µM and 10.3 µM compared to carmofur's IC50 of 13.0 µM. These results establish a structure-activity relationship between the composition of carmofur's side chain and its biological activity.

Edward Njoo

Benchtop 19F NMR spectroscopy enables catalyst optimization for the preparation of celecoxib and mavacoxib, 3-(trifluoromethyl) pyrazolyl benzenesulfonamides non-steroidal anti-inflammatory drugs (NSAIDs)

Anushree Marimuthu • Allen Ke • Arjun Akula • Andrew Chyu • Divya Gottumukala • Indalina Chan • Ishan Patel • Thomas Lavery

The abundance of fluorinated motifs in medicinal chemistry, many of which possess unique metabolic stability, have attractive scaffolds and serve as potent therapeutic agents. Of these, eighteen FDA-approved organofluorines have been classified as non-steroidal anti-inflammatory drugs (NSAIDs). Fluorinated COX-2 inhibitors celecoxib and mavacoxib 3-(trifluoromethyl) pyrazolyl benzenesulfonamides can be prepared through condensation of 4-sulfonamidephenylhydrazine and a trifluoromethyl diketone, for which quantitative kinetic analysis has yet to be established. With real-time benchtop 19F NMR spectroscopy, we are able to determine the absolute kinetics of pyrazole cyclo-condensation as well as the identification and quantification of transient intermediates which are otherwise unobservable. With this workflow in hand, we screened Lewis and Brønsted acids in order to enable catalytic preparation of celecoxib and mavacoxib; remarkably, we find that several organic sulfonic acids in methanol afford the most rapid conversion of starting material, outcompeting a panel of transition and rare-earth metal catalysts under certain solvent conditions (this trend is reversed in under other solvent conditions). Having identified optimized solvent and catalyst conditions in this reaction, we demonstrate that this is scalable to a gram-scale preparation of these two and other 3-(trifluoromethyl) pyrazolyl benzenesulfonamides. Ultimately, we envision future applications of this workflow in the synthesis of other fluorinated scaffolds, which are highly desirable motifs in a variety of medicinal chemistry settings.

Edward Njoo

Reactivity-informed Synthesis and Tumor-suppressing Activity of a Targeted Library of Close-in C19 Andrographolide Analogs

Alice Tao • Carina Zhou • Yilin Fang

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 anti-cancer therapeutic and is known to function putatively through the inhibition of Nf-kB, a transcription factor that modulates tumor survival and metastasis. Previous studies have found that modification at C19 leads to the Wnt/𝜷-catenin signaling pathway to be the andrographolide analog's primary mode of action. With this in mind, we synthesized a library of andrographolide analogs by protecting the C19 hydroxyl with large, hydrophobic silyl and trityl ethers, including the triisopropylsilyl, triphenylmethyl, and tert-butyldimethylsilyl ethers, to interrogate the role of C19 in the biological mechanism of action of andrographolide. We first evaluated the cytotoxicity of these analogs through MTT and cell migration assays. Quantitative polymerase chain reactions was then used to compare andrographolide and the analog's regulation of expression of MMP-7 and survivin, direct transcriptional outcomes from the activation of 𝜷-catenin. We performed western blots to assess the relative levels of active versus phosphorylated 𝜷-catenin. The relative gene expression of MMP-7 and survivin analyzed through q-PCR demonstrates down-regulation of such genes. Additionally, data from MTT assays revealed that all of our C19 analogs in HCT-116 cells showed greater cytotoxicity compared to that of andrographolide.

Harman Brah

Analogs of Irbesartan as GDP-Bound K-RAS Inhibitors for Targeted Cancer Therapy

Gargi Sharma • Aishwaryalakshmi Saravanan

KRAS (Kirsten Rat Sarcoma Virus) is a proto-oncogene involved in cell proliferation and division. It is mutated in approximately one-fourth of all cancerous tumors. We aimed to block GDP-bound KRAS' function, which is the inactivated version of the protein, which may help prevent malignant transformation in precancerous states. We tested the binding affinity of various small molecules using UCSF Chimera and Autodock Vina. We hypothesized that the addition of electron acceptors to the molecules would increase hydrogen bonds, thereby increasing thermodynamic favorability. We formed this hypothesis after docking several small molecules to GDP-bound KRAS and noticing that molecules with favorable free-energy binding affinities usually formed more hydrogen bonds with amino acid residues. Most of the small molecules that we tested were Benzimidazole derivatives; certain derivatives, such as methiazole and fenbendazole, have been found to inhibit KRAS-mutant cells. Given the mortality benefit seen with some benzimidazole derivatives, such as the Angiotensin II receptor blockers, we also investigated if one such drug, Irbesartan, and its analogs had a strong affinity for KRAS. Herein, we report a novel modification to Irbesartan that resulted in an increase in silico binding affinity for KRAS.

Harman Brah

Assessing Eight Docking Softwares for Accuracy and Speed with a Diverse Panel of Drugs

Aditya Narasimha

In recent years, molecular docking has become an indispensable tool in the domain of computational drug discovery. This prominence has prompted a rapid expansion in the availability of both open-source and commercial docking platforms. Given this proliferation, there's a pressing need to critically assess the accuracy, efficiency, speed, and versatility of these software solutions. In this study, ten ligands and their respective receptors were subjected to molecular docking across eight distinct software tools. The resultant binding affinities were benchmarked against values derived from established biochemical methodologies. Considering its prevalent utilization in scientific literature, we postulated that AutoDock would outperform its counterparts across the evaluated parameters. Among the softwares assessed, ParDock delivered values most congruent with literature-derived binding affinities, trailed by DINC and AutoDock Vina. Incorporating metrics of speed and reproducibility across the ten ligands, AutoDock Vina emerged as the superior choice. A nuanced understanding of the strengths and shortcomings of each software can provide invaluable guidance for researchers, enabling more informed choices in their computational endeavors.

Larry McMahan

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

Akshatvir Singh • Eesha Gadekarla • Laasya Nukala • Riddhi Sharma • Rishav Saravanan • Shivansh Bansal • Tiffany Liu

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. Futurewise, 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.

Larry McMahan

Jet Optimization Using a Hybrid Multivariate Regression Model and Statistical Methods in Dimuon Collisions

Vishnu Srinivas • Krishna Chunduri

Heavy ion collisions, in this case, muons, result in jets and noise. Jets are crucial event-shaped observable objects that are used in high-energy particle physics, i.e., they are one of the many objective variables that we can measure to determine the properties of a collision. However, many ionic collisions result in large amounts of noise, taking away from the jets' energy, increasing the energy lost, and thus reducing the efficiency of collisions with heavy ions. Our focus is specialized to dimuon collisions and improving the efficiency of said collisions. The purpose of our study is to analyze the relationships between properties of muons in a dimuon collision to optimize conditions of dimuon collisions and minimize the noise lost. We used principles of Newtonian mechanics and their use cases at the quantum level, allowing us to further analyze and conjecture different models and equations. We utilized tools such as Python algorithms as well as linear and polynomial regression models with tools such as sci-kit Learn, NumPy, and Pandas. Our hypothesis: the invariant mass, the energy, and the resultant momentum vector play a large role in determining the noise in a dimuon collision. If we constrain these inputs optimally, there will be scenarios in which the noise of the heavy-ion collision is minimized.

Michael Amadi

Developing a Protein-miRNA Complex as A Non-invasive Liver Cancer Biosensor

Gayathri Nair • Aayushi Agarwal • Ananya Miriyala • Ayaan Ahmed • Clytie Huey • Emilia Lee • Prital Jariwala • Sophia Walker

MicroRNA 122 (miR-122) is a liver-specific small RNA that regulates cellular differentiation of hepatocytes and tumor suppression in Hepatocellular Carcinoma (HCC). The RISC complex is composed of RNAse III endonuclease DICER1, Transactivation Response RNA binding protein (TRBP), and Argonaute 2, which bind to miRNA. Detecting miR-122 non-invasively and affordably in the early stages of HCC is challenging. To overcome this, we are developing a modified RISC complex that binds to miR-122, enabling its measurement and analysis in patients to determine the presence of HCC. By employing synthetic biology and special chemistries, biological components of the RISC complex and synthetic premature and mature miRNAs. Essential RISC complex proteins were synthetically produced using E. coli strains DH5α and BL21 (DE3). The DICER1 (responsible for cleaving premature microRNA) protein's endonuclease activity was then evaluated by incubating it with the synthetic premature microRNA labeled with 5' fluorescein and 3' Black Hole Quencher. Various temperatures were tested to identify the optimal temperature for DICER1's cleavage activity. The results showed that incubating DICER1 with premature microRNA led to fluorescence, primarily at 4℃, confirming DICER1's role in cleaving premature miRNA and the optimal temperature for DICER1 activity. These results will enable further investigation into the role of DICER1 in facilitating Argonaute 2 binding to miR-122. Our next steps involve evaluating the rest of the RISC proteins to understand their functions and potential applications in the development of the non-invasive biosensor.

Michael Amadi

Designing and developing a recombinant mRNA vaccine against the Nipah virus (NiV) by targeting fusion protein binding between mammalian and viral glycoproteins

Rajesh Veera • Caleb Yu • Anishka Duvvuri • Jayani Mamidi • Shreya Ray • Erin Law

The Nipah Virus (NiV) is a single-strand antisense RNA virus that relies on the specific binding of receptor proteins to increase its transendothelial migration across mammalian cells. The binding mechanism of NiV attachment protein G with EphrinB2 and Ephrin B3 activates F-mediated fusion, resulting in infection of the cell. However, current limitations in drug discovery efforts against NiV focus on the use of Remdesivir and Ribavirin and need to take into account the virus's innate ability to mutate, consequently rendering most organic small molecule approaches ineffective as long-term prophylactic agents. Here, we show the potential usefulness of a messenger RNA (mRNA) vaccine library as a stand-alone or combinatorial therapeutic agent along with FDA-approved drugs. Using in silico and in vitro models, recombinant NiV vaccine libraries were synthesized to replicate mutated receptor attachment glycoproteins. By using a codon optimization model, we evolve the NiV attachment proteins in an evolutionary congruent fashion. These candidate mRNA vaccines were encapsulated in lipid nanoparticles (LNPs) and transfected into mammalian cells to assess the efficiency of transfection and viability of the vaccine. Our early-stage mRNA vaccine library presents great potential since it is well-substantiated that these glycoproteins are required for viral entry. Similar to SARS-CoV-2, NiV exhibits rapid community transmission, however, with a higher mortality rate of 40-75%. This study proposes the effectiveness of a highly scalable and potentially low-cost mRNA vaccine library to circumvent the consequences of a NiV outbreak.

Michael Amadi

Development of synthetic aptamers for use as low-cost PLD1 inhibitors.

Anushka Sinha • Kavya Datt • Mansha Gupta • Izna Khanna • Divya Gill

PLD is a gene when expressed, produces two isoforms, PDL1 and PLD2. The PLD1 enzyme breaks down phosphatidylcholine into phosphatidic acid (PA) and choline via hydrolysis. PA binds to the active site of mTOR, initiating a phosphorylation cascade. Different kinases use ATP to amplify the signal that mTOR started, and mitosis begins once the signal reaches the centriole. PLD1 activity is relatively low in mammals and is transiently stimulated upon activation because of the G-Protein Coupled Receptor at the cell surface. GPCRs are involved in mitotic stimulations and stimulate mitosis, allowing cells to enter interphase. Interphase consists of the G1 and S phases responsible for cancer cell proliferation. The cell that initially had the overexpression of PLD1 due to a carcinogenic mutation will continuously conduct mitosis, as the phosphatidic acid will repeatedly trigger mTOR activation. Due to excessive amounts of phosphatidic acid the mTOR receptor constantly signals to the cell to start mitosis through DAG kinases, amplifying inflammatory anti-apoptotic functions, creating a roadblock for chemotherapy. We've developed an in silico method using SELEX to develop DNA or RNA synthetic aptamers with special chemistries to act as competitive PLD1 inhibitors to terminate the PLD1 signaling pathway, reducing cell proliferation and allowing patients to continue chemotherapy. Researchers have successfully produced PLD1, grown it, and performed bacterial transformations. The upcoming progress entails the use of SELEX, analysis of binding and enzyme kinetics, stability (half-life), and potential in silico-directed evolution analysis of PLD1.

Phil Mui

Transformers or Traditional? An In-Depth Analysis of Text-Based and Classification Machine Learning Models in Automating Small Business Loan Approvals

Aswin Surya • Aryan Madan • Aiden Li • Vineet Rao • Soham Vankudre • Tiffany Zhang • Karthik Subramanian

This study addresses the challenge of small business financing, highlighting the persistent issue of loan rejections despite good credit scores. The problem is often attributed to traditional bank loan approval processes failing to consider crucial financial factors. To address this, the research employs a combination of traditional machine learning and transformer-based methods to detect potential biases in loan approvals. It focuses on factors like loan term length, employee count, and other small business features. We applied XGBoost, CatBoost, Logistic Regression, and Random Forests, along with SHapley Additive exPlanation (SHAP) values, to assess the impact of different business features on loan approvals. The primary dataset comes from the U.S. Small Business Administration (SBA) and undergoes thorough preprocessing for model effectiveness. Additionally, the study compares human-driven preprocessing and model training with a GPT-4-based training pipeline using the Noteable plugin, including 0-shot, 1-shot, and few-shot learning. Surprisingly, the results reveal that the 0-shot and 1-shot learning models outperform human-curated models, while the few-shot learning model lags behind. The study underscores the potential of transformer architectures in automating the loan approval process and suggests future research directions, including exploring other Large Language Models (LLMs) beyond GPT and investigating ensemble techniques to enhance the reliability and applicability of the findings.

Phil Mui

Analyzing Factors Influencing Political Polarity in US Colleges

Ronit Kapoor • Hiresh Poosarla • Theodore Mui • Siqi Liu • Rishabh Garg • Anishka Vissamsetty • Maryam Solaiman

This study investigates political polarity trends across U.S. colleges. To compare across different colleges, we first identify online student newspaper opinion writing as the common type of data that we can collect. To collect comparable data, we create custom web crawlers to download and curate text corpora of college newspaper opinion writings for a diverse set of 32 colleges from 2010 to 2022. By applying natural language processing (NLP) to preprocess the text corpora, as well as using the Bi-Partisan Press' Polarity API, we measure political leanings of every opinion writing for each college over the dozen years. We then perform statistical analyses for each year of each college against a variety of demographic and institutional attributes. Our analysis confirms the existence of liberal bias in these college newspapers. In terms of factors influencing political polarity, we found a strong positive correlation between the voting trends of the surrounding county and the polarity of colleges within it. Among the colleges we analyzed, we also found a slight positive correlation between liberal bias and the ratio of males to females. Among the schools in our study, private liberal arts schools had the greatest trend toward more liberal views from 2010 to 2022. Finally, our analyses found a greater shift toward liberal ideologies for colleges in suburban settings over those in urban and rural areas.

Phil Mui

Classifying Technical Manuscripts via a Novel Transformer-Based Model

Aarush Gupta • Aryan Singhal • Niranjana Sankar • Vineet Rao

In this work, we introduce PRISM (Proprietary Rights Identification with Semantic Meaning parsing), a novel transformer-based classifier designed to identify technical manuscripts, particularly patents sourced from the USPTO (United States Patent and Trademark Office). Our model aims to accurately discern between various data sources containing intellectual property in an effort to achieve a comprehensive understanding of technical and creative works. We collect and organize an exhaustive dataset containing patents, political and tech-related news articles, and fictional content. PRISM utilizes the transformer encoder architecture by employing positional embedding layers, a self-attention layer, and a fully connected layer to classify a text as a patent or not. For developing PRISM, we leveraged the PyTorch and TensorFlow libraries, as they offer the essential components for tasks in deep learning, including transformer construction. For evaluation, we source patents from fields that differ from the ones used in our dataset, demonstrating the model's performance on unseen domains. Furthermore, we compare it against other deep learning architectures such as LSTMs and CNNs for further evaluation. Our approach is designed to outperform existing classification methods, ultimately providing PRISM as a precise classification system to differentiate technical manuscripts and offering insights into potential future applications of this technology.

Phil Mui

Mitigating Political Bias in Large Language Models Using Chain Of Thought Prompting Techniques

Hiresh Poosarla • Avni Goyal

Recent advancements in AI, particularly Large Language Models (LLM), highlight the challenges in aligning these systems' behaviors to be consistent with human values, intentions, and expectations. With our research, we are exploring the political biases inherent in LLMs. We apply various prompting techniques such as Chain of Thoughts (CoT) and study their effectiveness in moderating the amount of bias in ChatGPT. Our preliminary results are showing how proper prompting techniques are necessary to reduce LLM biases. At the same time, misleading prompting could in fact lead to increase in LLM biases.

Phil Mui

An Overview on the Performance of Reasoning Agents in Large Language Models

Maryam Solaiman • Ethan Shen • Ryan Li • Harish Senthilkumar • Rishi Gupta • Tanusha Tamijet • Vikram Subramanian

The recent rise of Large Language Models (LLMs), which are able to generate human-like text, has put a large amount of attention onto AI and its potential uses. However, most LLMs are limited to a one-dimensional/left-to-right method of decision-making that can impede their performance in tasks that require accurate foresight and reference to previous decisions to execute. We hypothesize that various types of LLM reasoning agents have different strengths and weaknesses that allow for different strategic use cases. In our research, we hope to determine the specific use cases and strengths of various reasoning agents, which will allow for the creation of LLMs tailored towards certain tasks with the use of such agents. With the help of reasoning agents, such as symbolic, arithmetic, and chain-of-thought reasoning, LLMs adopt a greater understanding of the context given to them and use a multi-step approach to adequately solve problems. Existing challenges in evaluating reasoning agents within LLMs include issues such as dataset biases and the potential brittleness of the models. These challenges, combined with the ethical concerns surrounding the reasoning agents such as their susceptibility to amplifying biases within a response, offer a rich research area. Using a quantitative analysis of several reasoning agents within a controlled environment, we apply diverse multi-modal and iterative reasoning techniques. Through this analysis, we explore the strengths and weaknesses of these reasoning techniques, resulting in a better understanding of the reasoning capabilities to be applied to real-world scenarios and products.

Robert Downing

Novel Structural Classification of DNA Binding Enzymes in Homo sapiens through the Geometry of the DNA Strand

Harsha Samavedam • Sweekrit Bhatnagar • Samarth Prajapati • Krishiv Aggarwal • Rionn Tuscano • Neetu Mathews • Sourodeep Deb • Gaurwik Paul

DNA binding enzymes play pivotal roles in driving fundamental biological processes, such as gene regulation, DNA repair, and replication. Understanding their structural characteristics is paramount for elucidating the mechanisms, interactions, and regulatory functions. In this study, we present a novel structural classification method for DNA binding enzymes in Homo sapiens, with a focus on the geometric aspects of the DNA strand, such as the macro-curvature, radius, and compression. By pinpointing and analyzing the phosphorus backbone of the DNA strand, we delved deeply into the common structural attributes shared across DNA strands when interacting with different enzymes, and we have determined varying structural features of each class of DNA-bound proteins. After collecting the essential structural data from the DNA strand, we are implementing multiple supervised learning methodologies. From the multiple models, we will determine the most accurate model to precisely map the correlation of specific structural features present in the DNA strand to the family of the bound protein. These findings can expand our comprehension of proteins and their functions, holding promise for applications in various biomedical domains, including drug development and pharmacotherapy.

Robert Downing

Key Molecular Descriptors Distinguishing Between Synthetic and Natural Products

Sweekrit Bhatnagar

The classification of natural products (NPs) from synthetic molecules (SMs) through machine learning techniques creates knowledge of differentiating features and therefore an impetus for possible research in natural product-based drug design. Natural products generally have a higher chemical diversity and biochemical specificity among other properties, making them favorable as lead structures for drug discovery and differentiating them from synthetic molecules. Here, we propose a machine-learning approach with the PaDEL descriptor software to develop a classification method to differentiate NPs and SMs with a variety of molecular features. An ensemble of supervised learning algorithms, including Logistic Regression, Naive Bayes, Random Forests, and Decision Trees, were tested to obtain the optimal feature importance amongst the molecular descriptors and highest accuracy. The experimental accuracy of the best-performing machine learning method outlined in this paper, Random Forests, reached an 89.19% accuracy, comparable with previous models performing the same classification. Identification and classification of distinguishable properties of natural products and synthetic compounds allows for a better understanding of available chemical data and better incorporation of such properties in small molecule drug discovery.

Robert Downing

Determining Trends in Key Pharmacokinetic Properties of FDA-Approved Therapeutic Drugs

Harsha Samavedam • Sweekrit Bhatnagar • Samarth Prajapati • Krishiv Aggarwal • Avirral Agarwal • Shrey Birmiwal • Tejal Malpeddi

Pharmacokinetics (PK) is the study of a drug and its metabolite kinetics in the body. Specifically, it refers to the evaluation of a drug and its metabolites in various areas of the body to determine its efficacy in administering treatment and prevent failures in drug design. At the cornerstone of PK analysis are ADME properties, standing for absorption, distribution, metabolism, and excretion, four core properties that define a drug's efficacy. Meanwhile, recent advances in deep learning have expanded the role of predictive ADME in medicinal chemistry, giving rise to various PK predictors to analyze the properties of a given drug. This paper aims to compare various web-based PK predictors and use them to determine trends in FDA-approved drugs' active ingredients and use a weighted-score analysis to assign a pharmacokinetic score to each molecule. We aim to provide conclusions on ideal PK properties for drugs targeting FDA approval as well as provide a standard for comparison of web-based PK predictors for future work in the field.

Robert Downing

A Comparative Analysis of Recent Progress in Blind Docking Softwares for Protein-Ligand Interactions

Harsha Samavedam • Tejal Malpeddi • Eliane Juang • Rionn Tuscano • Kavya Venkatesan • Aristaa Bhardwaj • Cheryl Lam • Justin Huang

In this study, we consider the advantage of blind docking as an unbiased way to identify and predict protein structures, binding sites, and their receptors to conduct a comparison of these software. With a thorough comparison of various blind-docking software, we aim to identify the most efficient and accurate methodology for blind-docking. We determined the scope of the analysis and comparison as including accuracy, predictive capability, algorithmic methodologies used, and the diversity of the data. Regarding the examination of these factors in an assortment of blind-docking software, each application generates information about the docking of the ligand. Thus, we analyze variables with strong correlation to the protein structures, such as the scoring function with the affinity of the ligand to the binding pocket and the estimated fitness (kcal/mol). The intention of our study is to provide a comprehensive analysis and comparison of blind-docking software for use in research pertaining to drug discovery and protein structure identification.

Sahar Jahanikia

Unlocking Academic Success: Empowering Neurodivergent Students with AI Assist

Adhvaith Ravi • Anish Gupta • Ayush Pareek • Bhadra Rupesh • Daniel Huang • Shriya Bhamidipati • Vedha Vora

Today, Attention-Deficit Hyperactive Disorder in American schools goes unnoticed due to a lack of student-teacher-parent communication (Wolraich, M. L., et al., 2005). Through our platform, AI Assist, neurodivergent students can improve their education through tasks ranging from focusing to note-taking in school and at home. We aim to guide students' learning in an effective manner, utilizing student-parent-teacher communication. To design our platform, we used Flutter, Python, Flask, and a large language model. From funny jokes reducing the stress levels of the student to utilizing amazing pictures to help improve memory retention, our platform allows students to engage in creative activities, which in turn helps improve their performance in daily activities. The app also has built-in options to play white noise and classical music to reduce the student's hyperactivity levels. Furthermore, the platform includes a notetaking feature that utilizes Large Language Models (LLMs) to help the student recall information learned in the classroom. Using audio processing to record class lectures, students will have access to continuous streaming notes through utilizing few–shot chain-of-thought prompting and analogical reasoning in order to understand material effectively. Using our platform, neurodivergent students are given the opportunity to better themselves and their knowledge through our numerous tools. AI Assist aims to help students with ADHD enhance their learning by creating an inclusive learning environment for academic success.

Vasudha Salgotra

Evaluating Naringin and Quercetin as P-Glycoprotein Inhibitors for Drugs With Low Bioavailabilities

Tulika Sarkar • Abir Bhatia • Sharada Kittur • Reva Ukkadam

Poly-glycoproteins (P-gp) are transporter proteins that act as efflux pumps for a range of drugs. These proteins are expressed primarily in liver, kidney, brain, and intestine cells. P-glycoproteins may have a significant effect on the bioavailability of drugs because of its anti-absorption mechanism. Naringin and quercetin have been hypothesized to inhibit P-gp and limit the expulsion of drugs. Our research focuses on testing the absorption, distribution, metabolism, and excretion (ADME) properties of these selected compounds and evaluating their potential as bioenhancers. SwissADME, an in silico tool, was used to understand the molecules' physicochemical parameters, which have been further investigated with in vitro testing. Data from the lipophilicity assays performed on naringin and quercetin suggested moderate absorption of these compounds by the body. In addition, plasma protein binding assays showed the percentage of compounds available at the P-gp site. The efficacy of these compounds will also be evaluated by in vitro ATP transporter and Caco-2 cell assays. Testing the feasibility of novel P-gp inhibitors can provide valuable insights on the optimization of various drugs with low bioavailability.