In-house, best-in-class antibody engineering, from VHH discovery and multi-specific antibody engineering to translational validation
Retrofit leverages a state-of-the-art toolbox to yield optimized and functionally characterized multi-specific antibodies tailored to the original clinical unmet need they will address.
Data packages are built aiming towards maximum human translatability: in vivo models are rationally selected via systems/AI-driven reverse translation based on their relevance to the human disease, and mode-of-actions are determined using an array of complementary translational human models. To increase clinical translation, for each asset, an AI-driven digital pathology-based biomarker is developed to guide future clinical development and allow for biology-informed patient stratification.
Critically, this is not a linear, one-way pipeline: wet-lab outcomes from target validation, preclinical efficacy/safety studies, and biomarker characterization are systematically fed back to iteratively refine our data-lake interrogation, target mining, and logic-gating approaches with each asset. This lab-in-the-loop architecture means the platform's predictive accuracy increases over time.