Biological macromolecules interact to form molecular complexes, enabling them to perform their functions. Understanding the structure, interactions, and functions of these complexes is a central goal of structural biology. In recent years, advances in deep learning have led to remarkable breakthroughs in addressing key challenges in this field through data-driven approaches. However, macromolecules are inherently dynamic, and their conformational heterogeneity directly affects their structure and function. While accounting for this flexibility is crucial, most current data-driven methods focus on static structures, often leading to inaccurate predictions in cases where flexibility plays a key role. My research focuses on developing novel, dynamic-aware, data-driven approaches to improve the prediction of macromolecular interactions and functions, specifically targeting host-pathogen interactions to elucidate antimicrobial resistance. Additionally, I aim to leverage these models for the design of novel functional proteins to fight against pathogens.