Protein Solubility Prediction
Developing deep learning models to predict protein solubility directly from sequence and structural features — work that feeds into faster, cheaper decisions in protein engineering and drug development pipelines.
01 — About
From accounting ledgers to neural networks.
I'm a PhD candidate in Computer Science at Tufts University, where I also work as a Research and Teaching Assistant. My path here wasn't linear — I started in accounting, moved through project management, crash-data analysis, and health informatics, before landing on the question I care about most: how deep learning can help us understand protein behavior. Today my research focuses on protein solubility prediction and protein–protein interaction (PPI) modeling, using deep learning to turn biological complexity into something computable. More broadly, I'm interested in general machine learning and modeling — the methods and ideas that carry over across problems well beyond biology.
Deep Learning
Protein Solubility Prediction
Protein–Protein Interaction (PPI)
Health Informatics
Machine Learning
Python
Data Analysis
Research & Teaching
02 — Research Focus
What I'm working on.
Protein–Protein Interaction (PPI)
Building neural network approaches to model and predict PPIs, with a focus on architectures that generalize across unseen protein pairs and scale to large interaction datasets.
Antigen–Antibody Binding Affinity
Modeling antigen–antibody binding affinity with deep learning, aiming to predict interactions that are useful for antibody design and immunology research.
03 — Publications
Papers.
MiLaSol: Modeling Protein Solubility by Mixing Up Multiple Protein Language Models
Weiwei Lou, Mert Erden, Lenore Cowen. Bioinformatics Advances, 2026.
04 — Contact
Let's talk.
Reach out about research collaboration, job opportunities, or anything deep-learning-modeling.