This work presents an automated approach for analyzing DSF traces that incorporates prior knowledge of the molecule under study. It automates the analysis of DSF thermograms, enabling scalable, consistent interpretation of large datasets without manual intervention. By improving the accuracy of extracted thermodynamic parameters, it allows DSF to approach the reliability of higher-resolution techniques, expanding its utility in antibody characterization and developability assessment.

The development of effective anti-fibrotic biotherapeutics is limited by a lack of human-relevant analytical frameworks linking immune modulation to functional tissue outcomes. Here, we present an immune-competent, human iPSC-based analytical platform that quantitatively resolves macrophage–fibroblast crosstalk in cardiac fibrosis. By integrating high-resolution spectral flow cytometry with functional co-culture and conditioned media assays, we define macrophage phenotypic states and directly map these states to cardiac fibroblast activation and extracellular matrix responses. This platform enables mechanism-linked potency and differentiation assessment of macrophage-targeting antibodies, supporting target validation, candidate ranking, and potency assessment for next-generation anti-fibrotic biologics.

Antibody discovery is entering a new paradigm driven by the convergence of biology and machine intelligence. We present the MAb-Intelligence Triad, integrating natural antibody evolution from clonally expanded, somatically hypermutated B cell repertoires, human expertise in library design and directed evolution (Kling-EVOLVE), and AI models trained on large-scale paired VH–VL datasets. Using SARS-CoV-2 as a benchmark, we demonstrate prediction of improved affinity and cross-reactivity, with emerging applications in oncology and autoimmunity.