Based on physical principles such as quantum mechanics, molecular mechanics, and statistical mechanics, ID4Gibbs™ has developed a proprietary molecular force field XFF, along with the XPose binding mode prediction and XFEP affinity prediction processes. These models strongly support molecular design, evaluation, and optimization for FIC, BIC, FF, and other drug discovery projects with high accuracy, efficiency, and throughput.

  • XFF-Proprietary molecular force field with high accuracy
    • High accuracy

      Based on feedback from >50 real pipeline applications, multiple rounds of iterations have been carried out on the force field fitting algorithms and parameter adjustments.

    • Massive training data

      Millions of quantum chemistry calculations and thousands of molecular dynamics simulations completed with hundreds of millions of CPU hours

    • Comprehensive coverage of chemical space for drug-like molecules

      Diversified and targeted internal training sets

    • Cloud platform deployment

      Automatic parameter correction, customizable proprietary force field development for customized chemical space

    • Comparison on affinity prediction

      Comparison of MM Vs. QM conformational energy (>100,000 molecular sets)

  • Xpose conquers "undruggable targets"
    • High accuracy

      Binding pose prediction results achieve atomic-level precision (RMSD<1.0 Angstrom);            

    • Strong performance

      Trained and tested on project systems, representative of real-world project needs;

    • Flexible docking

      For systems where ligand binding induces conformational changes in amino acid side chains. It can accurately predict the binding mode of ligands, as well as their position and conformation after binding.

  • XFEP high-throughput affinity evaluation
    • High accuracy

      Error within 1 order of magnitude in affinity prediction

    • High efficiency

      Accurate prediction of 5,000-10,000 small molecules within one week

    • High-Throughput computational performance

      10~100 fold throughput compared to alternative commercial tools

    • Comprehensive applications at low cost

      Covering substantial practical scenarios in real-life drug design, e.g., drug resistance and selectivity prediction, PROTAC, covalent binding molecules, de-novo peptide and protein design, FBDD/SBDD, rational drug design and optimization, functional group screening, pose validation, etc.