Enhancing Conformational Sampling for Intrinsically Disordered and Ordered Proteins by Variational Autoencoder

Overview of VAE: Variational Autoencoder (VAE) generates similar conformations based on a given MD trajectory, thus can be used to enhance the sampling of the diverse conformation for intrinsically disordered proteins (IDPs) and structural proteins.

Dependencies

• Compatible with Python 3.x.
• Dependencies can be installed using the requirements.txt file.

Training Model

1. Install all the requirements by executing pip install -r requirements.txt.
2. Install required protein .pdb processing library biobox from this github repository.
3. Next execute python preprocess.py pdb split which aligns the conformations from input pdb and creates VAE-required dataset at certain split.
4. To start a training run:

python packed_vae.py pdb

where pdb denotes the filename of input MD trajectory.

5. To calculate the RMSD between generated conformations and original ones:

python rmsd_min.py pdb

K-cluster

To conduct cluster analysis on generated conformations, please use the scripts in k-cluster folder.

1. Cluster analysis requires MMTSB tool set in this research, install the tool set from this website.
2. execute list.pl to get file list for further analysis.
3. execute run.sh to conduct cluster analysis.

Chemical Shift Calculation

To calculate secondary chemical shift of generated conformations, please use the scriptes in chemical-shift folder.

1. Calculation requires SPARTA+ program from this website.
2. execute split_pdb.py to calculate chemical shift.
3. execute avg_cs.py to get average chemical shift of target conformation ensemble.