Legend Biotech is a global biotechnology company dedicated to treating and, one day, curing life-threatening diseases. Headquartered in Bridgewater, New Jersey, we are advancing the next generation of cell therapy through a portfolio of innovative technologies, including autologous, allogeneic, and in vivo cell therapy approaches. Our research spans chimeric antigen receptor T-cell (CAR-T), T-cell receptor (TCR-T), natural killer (NK) cell-based therapies, and other emerging platforms designed to transform the treatment of serious diseases. From our global network of research and development sites, we are committed to discovering and developing safe, effective, and cutting-edge therapies for patients worldwide.

Legend Biotech entered into a global collaboration agreement with Johnson & Johnson to jointly develop and commercialize CARVYKTI® (ciltacabtagene autoleucel; cilta-cel). Through this strategic partnership, we combine complementary expertise and capabilities to advance innovative immunotherapies for patients living with multiple myeloma.

Legend Biotech is seeking a Principal Scientist, Bioinformatics as part of the Research & Development team based in Somerset, NJ.

Role Overview

Legend Biotech is at the forefront of developing innovative cell-based therapies. We are seeking a highly motivated and innovative Principal Scientist of Bioinformatics with focus on AI for Drug Discovery & Protein Engineering to join our multidisciplinary R&D team. In this role, you will be a key driver in development of generative AI models and agentic AI systems to discover novel disease targets.

Concurrently, you will apply geometric deep learning and structural biology principles to design and refine high-affinity binders (such as scFvs, VHHs, or novel scaffolds) specifically tailored for CAR-T cell therapies.

Key Responsibilities

AI-driven target discovery & agentic architecture.

Architect and deploy agentic AI systems to mine multi-omic datasets, clinical literature, and biological pathways for novel, tumor-specific drug targets; Build predictive frameworks using transcriptomics and other datasets to evaluate target draggability and potential off-tumor toxicity.

Generative AI & model development

Develop/adapt generative models (e.g., protein language models, diffusion models, and inverse folding algorithms) to design novel, highly specific binders for target tumor antigens; Model immune interfaces using advanced geometric deep learning to predict paratope-epitope interactions, ensuring high specificity and minimizing off-tumor toxicity.

Rational design, refinement & developability engineering

Inspect AI-designed binders to fix physical anomalies, optimize hydrophobic packing and hydrogen bonding networks, and resolve steric clashes or backbone strain; Tune binder biophysical properties to achieve optimal CAR-T signaling and prevent T-cell exhaustion.

Requirements

Ph.D. in Computational Biology, Computer Science, Structural Biology, or Biophysics.
3+ years of post-PhD experience (or 6+ years for MS/BS) in biotech/pharmaceutical industry or advanced research setting focusing on computational drug discovery, target identification, and antibody/binder design.
Expert programming proficiency and deep learning frameworks in Python and primarily PyTorch.
Experience building generative AI model, agentic AI systems utilizing LLM orchestration frameworks to query biomedical databases and literature.
Deep familiarity with structural biology software including AlphaFold, ESMFold, RFdiffusion, ProteinMPNN, BindCraft, and FoldFlow.
Proficiency in physics-based simulation and structural optimization/visualization tools (e.g., GROMACS, AMBER, OpenMM, PyMOL).
Solid understanding of structural immunology; Expertise in interface biophysics, including the ability to calculate interface binding energies (ΔΔG), evaluate electrostatic surface potentials.
Familiarity with foundational multi-omics platform datasets and target evaluation databases (e.g., TCGA, Open Targets, Human Protein Atlas, DepMap).