Role: 

Recent advances in computational sciences and machine learning (ML) have opened new avenues to discover and further optimize therapeutic antibodies.  

IGI, Inc. is a global, clinical-stage biotechnology company focused on developing innovative biologics in oncology. Headquartered in New York, NY, with Research site in Lausanne, IGI is advancing a robust pipeline of novel, first-in-class MultispecificsTM aimed at addressing complex diseases and treating patients holistically. Powered by its proprietary BEAT® technology platform, IGI is committed to delivering breakthrough curative therapies to improve and extend the lives of patients battling hematological malignancies and solid tumors.  

We are now seeking a creative and accomplished computational scientist leveraging machine learning as well as protein structure modelling to accelerate our antibody discovery platform. You’re enthusiastic about how massive antibody sequence and structural data can help build models to predict and improve antibody properties? You will be applying already built, generative machine learning models such as RoseTTAFold Diffusion or AlphaFold3, and be developing own models trained with our internal database and next generation sequencing (NGS) data from antibody discovery campaigns. You will be joining the protein engineering team as senior investigator and will be working in the interdisciplinary antibody discovery and engineering department with an experienced principal computational scientist. Your technical developments will have a tangible impact in speeding up and increasing the quality of our therapeutic antibody candidates. 

Key responsibilities: 

• Leverage AI/ML methods to enhance and accelerate the IGI antibody discovery and optimization engine, with particular emphasis on antibody affinity maturation. 

• Antibody sequencing data (including NGS) handling. Data extraction and application of available or internally developed ML models to help identify optimal candidates and remove antibody sequence liabilities based on sequence family relationships early in the discovery process. 

• Building an NGS database across antibody discovery campaigns to allow for ML model refinement and identification of predicted unspecific antibodies. 

• Interaction with experimentalists to further improve and refine built models using supervised machine learning feedback loops.  

• Support protein structure modelling including molecular dynamics, protein docking, computational mutagenesis and free energy perturbation (FEP) calculations. 

• Deploy existing, generative ML models such as RoseTTAFold Diffusion, AlphaFold3 to antibody discovery and engineering. 

• Deploy existing, discriminative ML models to assess the humanness, nativeness and developability of our antibody candidates. 

• In a cross-functional collaboration, contribute to our internal antibody database architecture and interface, allowing for development and training of ML models. 

• Collaborate closely with cross-functional teams including IT, protein engineering, lead identification, lead optimization to help curating their data and integrating it in our proprietarydatabase. 

• Contribute to scientific innovation through internal presentations, patents, external publications and presentations at scientific venues. 

Skills, knowledge and expertise: 

• PhD in machine learning/ computer sciences or related discipline preferentially with focus on biological applications such as protein engineering. 

• Proven track record in the independent development, training, application and evaluation of generative and optionally also discriminative ML models.  

• Strong competency in Python, and optionally familiarity with PyTorch or other relevant Python libraries, experience with modern software engineering best practices, preferentially applied to protein/antibody engineering. 

• Familiarity with the current state-of-the-art in ML-driven antibody engineering. 

• Experiences with protein molecular dynamics, protein docking, computational mutagenesis and optionally also free energy perturbation (FEP). 

• Excellent communication skills with proficiency in English. 

• Sufficient biotechnology domain knowledge to interact with interdisciplinary teams and in-particular familiarity with antibody structures, domains and sequence compositions including CDRs is a plus. 

• Scientific track record, first-author publications in journals or at conferences, preferred. 

• Curious mind enjoying a fast-paced environment and executing cross-functionally. 

• Nice-to-haves are experience with Rosetta, protein molecular dynamics simulation, de novo design, NGS data, Bayesian optimization, familiarity with antibody biology and drug development.