Working with AlphaFold2, RoseTTAFold and other predicted models

You can use the predicted models from AlphaFold and other prediction software in Phenix. These models can be very helpful in structure determination because the models can be very accurate over much of their length and the models come with accuracy estimates that allow removal of poorly-predicted regions.

How to use predicted models in Phenix

Phenix has tools for carrying out AlphaFold structure prediction, processing the resulting models, and for using these models in X-ray or cryo-EM structure determination. An in-depth overview with video tutorials describes these steps in detail. Here is a summary of the tools in Phenix.

You can use PredictModel to access the Phenix AlphaFold server and predict the structure of a protein chain.

To process predicted models, removing low-confidence regions and converting confidence estimates to B-values, you can use ProcessPredictedModel .

The PredictAndBuild tool will allow you to incorporate predicted models in the structure solution workflow. PredictAndBuild generates predicted models and uses them to solve an X-ray structure by MR or to interpret a cryo-EM map. The tool then carries out iterative model rebuilding and prediction to improve the models. The iterative procedure allows creating more accurate predicted models than can be obtained with a simple prediction. PredictAndBuild can automatically generate a fairly accurate model starting from just a sequence file and either cryo-EM half-maps or X-ray data. Additionally, it provides morphed versions of unrefined predicted models that can be useful as reference models for refinement.

Phenix resources for using predicted models: