I used AlphaFold to predict the 3D structure of a protein target that is not available in PDB. However, the prediction has low confidence values (low pLDDT scores). Should I proceed with molecular docking using this structure, or should I refine it first? If refinement is necessary, what are the best approaches to improve its reliability for drug design? Any recommendations or relevant research papers would be greatly appreciated!
If AlphaFold predicts a protein structure with low pLDDT scores, it suggests low confidence in parts of the structure. Here’s what you can do:
If critical regions are affected, you can avoid docking with a low-confidence structure and consider refinement or caution in your docking approach.
it means that your model has low confidence
try to Evaluate the pLDDT Scores
as there is this limit which suggests thet the pLDDT < 50: These regions are highly unreliable and shouldn’t be used for docking or further analysis. pLDDT 50–70: These regions are moderately reliable but might need refinement before use. pLDDT > 70: These regions are generally trustworthy and can be used for docking or other studies.
or avoid performing molecular docking on such low confidence protein model try to develop a new one, hope it helps ;]
- Badges
- Science topic
- Similar topics
- Linguistics
- Philology
- Classical Philology
- Manuscripts