CLI Workflow¶
In this short example, we will:
Generate an ensemble of 10 conformations for Histatin 5 (Hst5), an intrinsically disordered peptide with 24 residues;
Convert the generated ensemble to trajectory format;
Analyze the generated ensemble by calculating structural properties;
Reweight the generated ensemble using experimental SAXS data, and re-calculate structural properties using the optimal set of conformer weights.
1. Generate a protein conformational ensemble¶
To generate a conformational ensemble for your protein of interest, you simply need to provide Ensemblify with your parameters file.
For this example, we will use a parameters file created for Histatin5, a small intrinsically disordered peptide that can normally be found in saliva, where it acts as a defense against fungal infections.
Build your parameters file using the provided template, or with the aid of the more user-friendly HTML form.
In this example, we are using a YAML parameters file with the following information:
{'job_name': 'Hst5', 'sequence': 'DSHAKRHHGYKRKFHEKHHSHRGY', 'size': 10, 'databases': {'coil': '<path_to_database>'}, 'targets': {'A': [['MC', [1, 24], 'coil', 'TRIPEPTIDE']]}, 'output_path': '.'}
If you are following along with the example, please remember to update the <path_to_database> with the location on your machine of the database you want to sample from.
Generate an ensemble via the Ensemblify CLI:
Assuming you have an Ensemblify parameters file
Hst5_params.ymlin the current working directory, you can run:(ensemblify_env) $ ensemblify gen -p Hst5_params.yml Generating ensemble of 10 valid pdbs using 31 processor cores... Ensemblified!: 100%|██████████| 10/10 [00:07<00:00, 1.43valid_pdb/s] There are 10 valid pdbs, 14 were discarded ( 14 clashed | 0 violated constraints). Ensemble Generation Finished!
This command will create a directory in the current working directory named after the
job_nameparameter inHst5_params.yml. Assumingjob_nameis ‘Hst5’, a directory named ‘Hst5’ will be created and the ensemble will be located in./Hst5/ensemble/valid_pdbs.
2. Convert the generated ensemble to trajectory format¶
After generating an ensemble, you can convert your set of PDB files into XTC format, a compressed trajectory format from GROMACS. This allows for much more efficient storage of created ensembles with minimal loss of structural information. Additionally, we can take advantage of the many available methods for analyzing files in this format.
To do so, you need only provide the location of your stored ensemble and the directory where you want to store your trajectory, optionally defining a prefix identifier for the created file.
Along with the created trajectory, one of the structures of the generated ensemble is saved as a topology PDB file that contains atomic connectivity information, as this is not stored in the trajectory and is required when using the analysis module.
Assuming the current working directory is the previously created ‘Hst5’ directory:
(ensemblify_env) $ ensemblify con -e ./ensemble/valid_pdbs -t ./trajectory -i Hst5 Hst5 Trajectory creation complete! : 100%|██████████| 4/4 [00:00<00:00, 475.09step/s]
The created trajectory and topology files will be in
./trajectory/Hst5_trajectory.xtcand./trajectory/Hst5_top.pdb, respectively.
3. Analyze your generated ensemble¶
After creating your trajectory, you can use it to create an interactive analysis dashboard with different plots and figures which will aid you in the structural analysis of your protein using your created ensemble. To do this, specify the location of your trajectory and topology files and the output directory where you want to store your interactive dashboard, the individual figures, and the data used in its creation.
Assuming the current working directory is the previously created ‘Hst5’ directory:
(ensemblify_env) $ ensemblify ana -trj ./trajectory/Hst5_trajectory.xtc -top ./trajectory/Hst5_top.pdb -tid Hst5 -out ./analysis Analyzing Hst5 trajectory... Calculating ramachandran data for Hst5... Calculating contact matrix for Hst5... Calculating contact matrix...: 100%|██████████| 10/10 [00:00<00:00, 810.87it/s] Calculating distance matrix for Hst5... Calculating distance matrix... : 100%|██████████| 10/10 [00:00<00:00, 2737.62it/s] Calculating secondary structure assignment frequency matrix for Hst5... Calculating structural metrics data for Hst5... Calculating rg... Calculating eed... Calculating dmax... Creating Hst5 analysis figures... Building ['Hst5'] analysis dashboard... Ensemble analysis calculation has finished. Please consult the interactive analysis_dashboard.html figure.
You could now open your
analysis_dashboard.htmlfile in your browser and interpret your results.
4. Reweight your generated ensemble using experimental SAXS data¶
After generating an ensemble, you can use experimental SAXS data to reweight it. This will create an interactive reweighting dashboard with comparisons between your uniformly weighted and reweighted ensembles, both in regards to fitting to experimental data and the calculation of structural properties of your protein.
To do this, specify the location of your trajectory, topology and experimental SAXS data files and the output directory where you want to store your interactive dashboard, along with the individual figures and data used in its creation.
Assuming the current working directory is the previously created ‘Hst5’ directory, and the
Hst5_SAXS.datfile is one directory up:(ensemblify_env) $ ensemblify rew -trj ./trajectory/Hst5_trajectory.xtc -top ./trajectory/Hst5_top.pdb -tid Hst5 -exp ../Hst5_SAXS.dat -out ./reweighting Processing Hst5 experimental data file... Experimental errors on SAXS intensities have been corrected with BIFT using scale factor 1.0. Calculating Hst5 SAXS data... : 100%|██████████| 10/10 [00:00<00:00, 260.36it/s] Applying BME reweighting to Hst5 ensemble with theta values [1, 10, 20, 50, 75, 100, 200, 400, 750, 1000, 5000, 10000] ... Reweighting ensemble... : 100%|██████████| 12/12 [00:00<00:00, 93.66it/s] Please analyze the provided interactive figure (effective_frames_fit.html) and input the desired value(s) for the theta parameter. If more than one value, please separate them using a comma. Chosen theta value(s): 200. No contact matrix data was provided. Calculating contact matrix...: 100%|██████████| 10/10 [00:00<00:00, 910.24it/s] Calculating reweighted contact matrix...: 100%|██████████| 10/10 [00:00<00:00, 883.94it/s] No distance matrix data was provided. Calculating distance matrix... : 100%|██████████| 10/10 [00:00<00:00, 1757.51it/s] Calculating reweighted distance matrix... : 100%|██████████| 10/10 [00:00<00:00, 1744.50it/s] No secondary structure assignment frequency matrix data was provided. Calculating secondary structure assignment frequency matrix... Calculating reweighted secondary structure assignment frequency matrix... No structural metrics distributions data was provided. Calculating rg... Calculating eed... Calculating dmax... Creating Hst5 reweighted interactive figures... Building Hst5 reweighting dashboard... Ensemble reweighting has finished. Please refer to the interactive reweighting_dashboard.html figure for analysis.
You could now open your
reweighting_dashboard.htmlfile in your browser and interpret your results.