Research
Andrei Manolache develops representation learning methods for graphs, molecules, and proteins. His work includes learning approximate equivariance through explicit symmetry constraints and combining molecular connectivity with three-dimensional structure in self-supervised pretraining. He also investigates large-scale protein fold classification and pretraining, connecting methods for structured data with biological applications.
Selected publications
- Protein Fold Classification at Scale: Benchmarking and Pretraining 2026
- Learning the Neighborhood: Contrast-Free Multimodal Self-Supervised Molecular Graph Pretraining 2026
- Learning (Approximately) Equivariant Networks via Constrained Optimization 2025
Information reviewed 26 September 2026 · University profile