A tutorial to setup and run AlphaFold-Metainference simulations. The original work which explains the theory for AlphaFold-Metainference (AF-MI) can be found here.
The tutorial comprises the following sections:
A brief introduction to the theory and advantage in using AlphaFold-Metainference.
We will present the case study for this tutorial, the DNA/RNA processing protein TDP-43 which undergoes both functional and pathogennic aggregation in amyotrophic lateral sclerosis (ALS) and other neurodegenerative conditions. To enable a rapid structural ensemble generation on a laptop/workstation, AF-MI is coupled to the CALVADOS-2 coarse-grained forcefield.
We will discuss all the preliminary steps to set up AF-MI, from generating the AF2 distance-map restraints data, to preparing PLUMED inputes, and running energy minimization and production runs
Finally, we will discuss how to generate the unbiased structural ensemble of TDP-43, estimate free energy surfaces along collective variables (CVs) and predict ensemble averaged properties such as secondary structureand RMSF.
NB: This tutorial assumes that you know metadynamics theory and practice in PLUMED (if not, please visit PLUMED tutorial).
flowchart TB
A[PLUMED syntax] -.-> D[Theory];
B[Parallel Bias Metadynamics] -.-> D;
C[Metainference] -.-> D;
D --> E[Case study];
E --> F[Jupyter/Colab]
F --> G[Inputs]
G --> F[Jupyter/Colab]
F --> H[Post-processing]
click A "syntax" "This lesson teaches you the basic features of the PLUMED input syntax"
click B "PBMetaD" "This lesson teaches you how to perform and analyze Metadynamics simulations"
click C "Metainference" "This lesson teaches you how to perform and analyze Metadynamics simulations"
click D "theory.md" "AlphaFold-Metainference theory intro & overview"
click E "case_study.md" "Case study: TDP-43 monomer dyamics"
click F "AF-MIcolab" "AF-MI colab performing software installation, system preparation, AF-MI simulation and post-processing analysis"
click G "input.md" "Input: Protein sequence, AF-MI simulation data and parameters"
click H "postprocessing.md" "Structural ensemble generation and post-processing analysis"