Waste Minimized Modular Synthesis of Lactam-fused Oxygen Heterocycles as Potential Antifungal Agents Ryan Weekley, Keegan Gales Project Mentor(s): Timothy K. Beng, PhD Fungal diseases are pervasive and present problems for farmers and society at large. This causes billions of dollars in annual losses and threatens global food security. Thus, the need to develop new and better methods to combat this problem is growing with every passing day. Addressing this situation is a responsibility entrusted to synthetic organic chemists. In recent years, the Beng research group has sought to identify more selective and specific medications that can either prevent or treat several medical conditions. Our prior efforts identified some lactams (an amide in a cycle) as efficient antithrombotic and anti-cancer agents. Current literature shows that a few lactams are effective fungicides. Furthermore, oxygen-containing heterocycles (e.g., tetrahydropyrans (6-membered rings) and tetrahydrofurans (5-membered rings) make up the core structure of some existing fungicides. We reasoned that if one molecule bearing both the lactam and oxygen heterocycle can be prepared efficiently, this would undeniably expand the space for the discovery of new anti-fungal medicines. Herein, we report our progress toward the construction of a diverse library of lactam-fused oxygen heterocycles in a cost-effective, efficient, and waste minimized manner. Our methodology hinges on the use of a green solvent such as hexafluoroisopropanol (HFIP) and potentially an electrochemical cell. The structures of the synthesized compounds have been confirmed by routine techniques, including nuclear magnetic resonance (NMR). Through a collaborative effort, the ability of these novel compounds to combat fungal infections will be evaluated in the future. Presentation Type: Poster Presentation (May 21, 9:30am–3:00pm) Keywords: Potential antifungal lactams and heterocycles SOURCE Form ID: 143 Computer Science AI IDE: An Intelligent Web-Based Code Editor with Agentic Code Generation Orchlon Chinbat Project Mentor(s): Razvan Andonie, PhD This project presents AI IDE, a web-based integrated development environment that augments traditional code editing with an intelligent, multi-agent code generation pipeline. The system combines a Next.js frontend featuring a Monaco-based editor, file tree management, and an integrated terminal with a FastAPI backend powered by LangGraph for orchestrating specialized AI agents. The core of the system is a Retrieval-Augmented Generation (RAG) pipeline that indexes project code into a Supabase pgvector store, enabling context-aware code generation grounded in the user's actual codebase. Upon receiving a user query, a router agent classifies task complexity and delegates to a planner (for complex tasks) or directly to a code generator, followed by an automated code reviewer that iteratively refines output for up to three revision cycles. The platform supports multi-file project management with drag-and-drop uploads, real-time streaming of AI responses, in-browser code execution across over ten programming languages, and persistent project storage. By combining local LLM inference via Ollama with optional OpenAI integration, multilingual embedding models for semantic code retrieval, and an iterative generate-review loop, AI IDE provides a self-contained environment for AI-assisted software development, rapid prototyping, and educational use. Presentation Type: Oral Presentation (May 20, 9:30am–5:00pm) Keywords : Artificial Intelligence, Retrieval Augmented Generation, Coding, AI Agents, Lang graph SOURCE Form ID: 75
111
Made with FlippingBook interactive PDF creator