My grandfather passed away from cancer, and after speaking with 100+ biomedical researchers and professors, we realized how slow (decades), expensive ($1B+ for one drug), and inefficient modern drug discovery still is. While that infrastructure stays broken, millions of people die waiting for drugs that either don't exist yet or never make it through the pipeline.
AetherBio is an AI-powered computational preclinical engine that generates, refines, and simulates drug candidates across multi-organ digital twins before physical testing begins.
We built AetherBio by chaining a GraphSAGE GNN, VAE/diffusion molecular generator, QAOA/VQE Hamiltonian minimizer, Hidden Markov conformational statistical modeler, and a multi-organ digital twin solving two-compartment PK ODEs, Michaelis-Menten statistical kinetics, and tissue diffusion PDEs, all connected through a React frontend and Python/Qiskit backend.
Some of the biophysics statistical models are:
$$ v = \frac{V_{max}[S]}{K_m + [S]} $$
$$ \frac{dP_i(t)}{dt} = \sum_j \left( k_{ji}P_j(t) - k_{ij}P_i(t) \right) $$
The hardest challenge was making every simulation layer scientifically coherent while connecting them sequentially, ensuring the plasma concentration curve from the PK ODEs fed live into the toxicity accumulation model, the Markov stationary distribution informed the quantum energy landscape, and the Nernst-Planck PDE stayed numerically stable under Euler integration without blowing up at the boundary conditions.
We're proud of successfully building an end-to-end interactive platform capable of simulating therapeutic behavior digitally across the human body.
We learned how difficult biological system modeling is at scale and how important clear visualization is for communicating complex scientific ideas.
We plan to collaborate with researchers at UC Davis to improve biological simulation accuracy, integrate larger-scale molecular datasets, and explore quantum annealing workflows using platforms like D-Wave Systems to accelerate conformational optimization and therapeutic discovery.
Log in or sign up for Devpost to join the conversation.