DL structure tools introduce chiral errors such as swapped chiral centers, D-amino acids, and cis-amide bonds

Summary

Deep learning structure prediction tools introduce errors such as swapped chiral centers, steric clashes, and cis-amide bonds (1,2). Diffusion-based structure prediction methods are able to overcome this with guidance.

Details

Antibody-specific tools include AbLooper, DeepAb, IgFold, ImmuneBuilder, and MOE homology modeler. Only DeepAb and ImmuneBuilder did not introduce D-amino acids.

See also

• DL ligand docking methods generate unrealistic poses

1.

Abramson J, Adler J, Dunger J, Evans R, Green T, Pritzel A, et al. Accurate structure prediction of biomolecular interactions with AlphaFold 3. Nature. 2024;630(8016):493–500. Available from: https://doi.org/10.1038/s41586-024-07487-w

2.

Fernández-Quintero ML, Kokot J, Waibl F, Fischer A-LM, Quoika PK, Deane CM, et al. Challenges in antibody structure prediction. mAbs. 2023;15(1). Available from: https://doi.org/10.1080/19420862.2023.2175319