AI-POWERED PROTEIN STRUCTURE PREDICTION

  – Boltz is an AI-based, all-atom structure prediction algorithm that models proteins, DNA, RNA, and small molecules, including their interactions within biomolecular complexes. Use Boltz when you want to predict a protein structure in the context of binding partners, investigate protein–ligand interactions, or model complexes containing multiple types of molecules. Source: Boltz

  – AlphaFold2 is the faster ColabFold implementation of AlpaFold 2.3 (AlphaFold-Multimer), known for its ability to predict protein structures from amino acid sequences with high accuracy. Use AlphaFold2 when you want to predict the three-dimensional structure of an individual protein, particularly when you want to establish a baseline model for structural analysis or comparison with predictions from other algorithms.

  – OpenDDE is an all-atom, diffusion-based structure prediction algorithm that can model proteins, nucleic acids, small molecules, and their interactions. Use OpenDDE when you want to investigate biomolecular complexes, including protein–protein and antibody–antigen interactions, or explore predicted binding poses involving multiple molecular components.

  – ESMFold2 is a protein structure prediction algorithm based on protein language models that predicts structures directly from amino acid sequences without requiring a multiple sequence alignment (MSA). Use ESMFold2 when you want a rapid first-pass structure prediction, need to model a protein with limited available sequence homologs, or want to screen multiple sequences efficiently.

  – ColabDock is an AI-powered protein–protein docking method designed to predict how two proteins interact and orient themselves within a complex. Use ColabDock when you already have two protein sequences or structures and want to investigate their potential binding arrangement, such as when studying protein–protein interactions or antibody–antigen binding.