-
Notifications
You must be signed in to change notification settings - Fork 0
Tutorial 5.5 AlphaFold_Unmasked
LIT-AlphaFold allows to perform predictions of multimeric structures using the AlphaFold-Unmasked protocol. In this way inter-chain distances extracted from structural templates are used during the calculation to guide the structure prediction.
To achieve this it is necessary to give to AlphaFold a template structure containing all relevant chains, with at most one exception. LIT-AlphaFold is able to automatically identify a valid Multimer template between the structural templates selected for each monomeric unit, and it assign to each monomer a different chain template. In the current implementation there is no way to control which monomer will be associated to whch template. This can be overcome by perfomring template filtering for each unit before running the predictions.
To run calculations using AlphaFold-Unmasked
python $LITAF/run_multimer_jobs.py \
run=af_unmasked \
output_path=test_structure \
monomer_objects_dir=[../tutorial_1_2/protein_features_mmseqs2] \
input_file=[multimer.in.yaml] \REFERENCE: MIRABELLO, Claudio; et al. Unmasking AlphaFold to integrate experiments and predictions in multimeric complexes. Nature Communications, 2024, 15, 8724.doi: 10.1038/s41467-024-52951-w
- Introduction
- Tutorials
- Tutorial 0: Introduction and Environment variables
- Tutorial 1: Input generation
- Tutorial 2: Input customization
- Tutorial 2.5: Database query
- Tutorial 3: Monomer predictions
- Tutorial 3.5: Prediction parameters
- Tutorial 4: Multistate predictions
- Tutorial 4.5: Additional features
- Tutorial 5: Multimer predictions
- Tutorial 5.5: AlphaFold-unmasked