Diana Garnica Acevedo & Jillian Brejnik, Graduate Students, Kostal Group, GW Department of Chemistry
Friday, September 5, 202510:00 am - 11:00 am
Image
In Silico Models to Design Safer Small Molecules
Computational chemistry is a valuable tool in modern science, providing insights into chemical structure, reactivity, and interactions between molecules and their environment. The use of in silico descriptors combined with statistical methods enables the development of predictive models for small molecules, facilitating early-stage screening and decision-making prior to costly experimental studies. These approaches support the design of safer chemicals that not only comply with current regulations but also are less likely to accumulate in biological systems or cause unintended adverse effects in humans. Persistent compounds, particularly lipophilic ones, tend to bioaccumulate in storage sites such as adipose tissue where they can remain for long periods and increase the risk of toxicity. To address this gap, our work focuses on developing predictive models addressing two critical areas of concern: biodegradation and lipophilicity of compounds with molecular weights up to 1,000 g/mol. For Biodegradation, our approach is based on a binary decision-tree that delivers robust binary assessments of readily biodegradable (RB) versus non-RB compounds. This method integrates site-specific SARs that mimic enzymatic binding with a physics-led strategy leveraging density functional theory (DFT) to evaluate molecular responses to electron-density flux, a proxy for the redox processes occurring during Phase I/II transformations in vivo. For Lipophilicity, our approach is based on a Multiple Linear Regression (MLR) model that predicts the distribution coefficient (logD) at physiological pH by capturing how a compound partitions between water and octanol using Lennard-Jones and Coulomb interaction energies as descriptors. This strategy provides a physically grounded basis for modeling lipophilicity as they contribute to solute–solvent stabilization.
BIO
Diana earned her B.S. in Chemistry in her home country, Colombia, and later completed her M.S. in Chemistry at The American University, where she specialized in organic synthesis of small molecules. Since joining the Kostal Research Group in Spring 2024, she has been working on predictive modeling approaches to better understand small molecule behavior .
Use of In Silico Methods to Limit Nitrosamine Formation and Hazard in Cosmetics and Personal-Care Products
Most nitrosamines are classified as carcinogens that are formed in a reaction between a vulnerable amine and a nitrite compound. Although originally found in pharmaceuticals, nitrosamines can also form in cosmetic and personal care products, thus impacting a large group of consumers. Raw materials, preservatives, reaction conditions, and production habits can all be related to the presence/formation of nitrosating agents, such as the nitronium ion, nitrogen trioxide, or NOx in a cosmetic formulation. To prevent nitrosamine formation, the addition of a scavenger such as an antioxidant or amino acid can be added to a formulation. The purpose of the scavenger molecule is to compete with the amine for the nitrosating agent to produce an inactive form of nitric oxide, thus allowing the amine to remain intact with no impurity. In silico modeling using Density Functional Theory (DFT) can provide a cost-effective way to accurately screen potential transformations and compute the kinetics and thermodynamics of nitrosamine formation potential, including their transition states. Calculated energetics can then be leveraged to best understand the underlying structural drivers of nitrosamine formation. This project considers a range of secondary amines, nitrosating agents, and commonly used solvents, which have not been extensively studied. Since not all nitrosamines are the same, the mutagenic/carcinogenic potential of the impurity will be concurrently assessed using the regulator-approved Carcinogenic Potency Categorization Approach (CPCA). This is complemented by modeling the binding of nitrosamines to relevant Cytochrome P450 isoforms to assess their potential for metabolic activation and the resulting capacity for DNA alkylation. This in-silico approach aims to enhance the evaluation of both existing and future formulations, enabling concerns to be addressed at early stages of the design process to limit the adverse effects of cosmetics and personal care products on human health.
BIO
Jillian earned her B.S. in Chemistry and Environmental Science at Washington College on the Eastern Shore of Maryland. Here, she researched the impact of hazardous chemicals in cosmetics products on algae growth for her senior thesis. Since joining the Kostal Research Group in Fall 2023, she has been working on modeling nitrosamine impurity formation potential in cosmetic and personal care formulations.
