Scientific Publications
Peer-reviewed research from our team spanning AI, drug discovery, and computational biology.
Destructured Drug Discovery: How Sequence-Based AI Speeds and Expands the Search for New Therapeutics
American Chemical Society / C&EN white paper featuring Ainnocence's sequence-based AI platform — 80% cost reduction, billion-candidate virtual screening, and validation on SARS-CoV-2 broad-neutralizing antibodies.
AINN-P1: A Compact Sequence-Only Protein Language Model Achieves Competitive Fitness Prediction on ProteinGym
A 167M-parameter mLSTM-based protein foundation model leading ProteinGym on stability (0.625, +6% vs ProSST, +39% vs 100B xTrimoPGLM) and generalizing to unseen antibody programs (+0.15 AUC vs 650M ESM2). Roger Wang, Kevin Jin, Lurong Pan.
AI-designed, mutation resistant broad neutralizing antibodies against multiple SARS-CoV-2 strains
Demonstrating the power of sequence-based AI for therapeutic antibody design with unprecedented speed and accuracy.
CarbonAI: A Non-Docking Deep Learning Based Small Molecule Virtual Screening Platform
Novel approach to virtual screening achieving 10B compound capacity in hours.
2021Sequence-based Deep Learning Antibody Design for In Silico Affinity Maturation
Foundation work for SentinusAI's antibody optimization capabilities.
2021ParaVS: A Simple, Fast, Efficient Graph Neural Network Framework for Virtual Screening
Breakthrough in structure-based virtual screening efficiency.
2025DeepLigType: Predicting Ligand Types of Protein-Ligand Binding Sites Using Deep Learning
Advanced binding site prediction for drug discovery applications.