Growing ecosystem of deep learning methods for modeling protein$\unicode{x2013}$protein interactions
This is an incremental review paper that synthesizes existing deep learning approaches for researchers in computational biology and bioinformatics, without presenting new methods or results.
The paper reviews the growing ecosystem of deep learning methods for modeling protein-protein interactions, highlighting their use of experimental data and biophysical knowledge to tackle the challenge of diverse molecular recognition mechanisms in the proteome, with recent successes in representation learning, geometric deep learning, and generative modeling for tasks like predicting interactions and designing assemblies.
Numerous cellular functions rely on protein$\unicode{x2013}$protein interactions. Efforts to comprehensively characterize them remain challenged however by the diversity of molecular recognition mechanisms employed within the proteome. Deep learning has emerged as a promising approach for tackling this problem by exploiting both experimental data and basic biophysical knowledge about protein interactions. Here, we review the growing ecosystem of deep learning methods for modeling protein interactions, highlighting the diversity of these biophysically-informed models and their respective trade-offs. We discuss recent successes in using representation learning to capture complex features pertinent to predicting protein interactions and interaction sites, geometric deep learning to reason over protein structures and predict complex structures, and generative modeling to design de novo protein assemblies. We also outline some of the outstanding challenges and promising new directions. Opportunities abound to discover novel interactions, elucidate their physical mechanisms, and engineer binders to modulate their functions using deep learning and, ultimately, unravel how protein interactions orchestrate complex cellular behaviors.