AI/ML Scientist

At GenSci, we are committed to improving the health and well-being of children and women around the world. Established in 1997, Changchun GeneScience Pharmaceutical Co., Ltd. is a subsidiary of Changchun High-Tech Industries Group and a leading genetic engineering pharmaceutical company. We are Asia’s largest recombinant human growth hormone manufacturing base and provide a comprehensive portfolio of recombinant growth hormone products. Our teams work across pediatric and women’s health, adult endocrinology, dermatology, rheumatology and immunology, and oncology to develop innovative solutions for unmet medical needs. With nearly 30 years of experience, we integrate R&D, production, marketing, and patient services to serve patients worldwide.

Role Overview

Our AI Research Institute is expanding its core computational drug discovery team. We are looking for an exceptional AI/ML Scientist (AIDD/CADD) who excels in applying modern machine learning, foundation models, and structural biology approaches to novel drug design.

Depending on your background, you will take ownership of ONE of our core modality design tracks ( Proteins, Peptides,double resistance, or siRNA ) and drive the dry-wet loop integration to push therapeutic candidates into development

Education

• PhD or Master's in Computational Chemistry, Machine Learning, Cheminformatics, Molecular Modeling, Protein Engineering, Structural Biology, Biophysics, Medicinal Chemistry, Chemical Biology, Immunology, Organic Chemistry or related fields
• Fresh PhD graduates and outstanding Master's graduates welcome
• Dual background in life sciences + computer science is a strong plus

Experience

• 0–3+ years of relevant industry experience in AI/ML-driven drug design, computational chemistry, molecular modeling, computational protein design, biologic engineering, or small molecule discovery
• Experience with wet-lab metrics (antibody humanization, immunogenicity, developability; peptide synthesis & activity; oligonucleotide ADME) and ability to collaborate cross-functionally
• Full-process drug discovery experience (target → hit → lead → PCC) highly valued
• Library construction & screening experience is a plus

Core Competencies

AI / Machine Learning

• Machine Learning, Deep Learning, Generative AI
• Transformer Architecture, Protein/RNA Language Models (ESM-2, RNA-FM, antibody LMs)
• Fine-tuning, LoRA / PEFT, SFT, RLHF / Preference Optimization, Domain Adaptation
• Graph Learning, Diffusion / Flow-Based Models, Multimodal Learning
• Model Evaluation, Ablation Studies, Uncertainty Estimation, Model Calibration, Interpretability

CADD / Computational Chemistry

• Computer-Aided Drug Design (CADD), AI-Driven Drug Discovery (AIDD)
• Molecular Dynamics (MD), Free Energy Perturbation (FEP)
• Quantum Chemistry, Force Field Parameterization
• Protein-Ligand / Protein-Protein Interaction Modeling
• Docking, Virtual Screening, Structure-Based Drug Design
• De Novo Design, Cheminformatics, QSAR

Therapeutic Modalities

• Antibody Engineering: humanization, immunogenicity optimization, developability (thermal stability / PTM / viscosity / non-specific binding), affinity maturation, bispecifics / multispecifics, TCE design
• Peptide Engineering: secondary structure analysis, chemical modification (lipidation, PEGylation, D-amino acid, N-to-C blocking), cyclization, stability & penetration enhancement, SPPS
• Oligonucleotide / siRNA: mechanism of action, sequence design, off-target control, nucleic acid modifications (2'-OMe, 2'-F, 2'-MOE, LNA/cET, phosphorothioate, GalNAc), ADME properties, endosomal escape

Tools & Software

• Programming: Python, PyTorch, JAX, RDKit
• CADD Suites: Schrödinger, MOE, Discovery Studio, OpenEye
• AI/Structure Models: AlphaFold 2/3, Boltz-1/2, BoltzGen, RFdiffusion / RFdiffusion2, ProteinMPNN, LigandMPNN, Rosetta, WeMol
• ML Tools: DiffDock, Uni-Mol, Chemprop, DeepChem

Soft Skills

• Cross-functional collaboration & communication
• Project ownership & delivery mindset
• Scientific rigor in model validation
• Ability to translate computational insights into experimental hypotheses
• Fast learning & adaptability across therapeutic modalities