Manual

1. Collect PDB ensembles

First, you need to create an ensemble of PDB files. There are no restrictions on the source of the PDB files. You can run MD simulations and extract time frames. Alternatively, you can perform docking or folding.

Since the goal is to construct a pathway that connects a start and end conformations, you need to make sure that you have sampled your system densely enough. The more dense the sampling, the smoother the pathway can be.

All PDB files have to contain the same atoms. They must describe the same molecular system.

2. Add energy values to PDB files

SmoothT searches for low energy pathways. Since SmoothT does not calculate the energy itself, it relies on the energy values being written into every PDB file.

You can add a line anywhere in the file. The line has to contain an identifier, any string that is unique in the file! The last entry of that line has to be the energy or score of the given PDB file (separated by a space from previous entry).

There are no restrictions on the source of that score. However, it is advisable to use one kind of scoring or energy function for all models.

Example line:

'ENERGY_IDENTIFIER  ...  -12.34'

Some programs write the energy values by default into the PDB files. For example, all tools in the Rosetta suite add a block of scores to the end of each PDB file they create. The first lines of such a block are as follows:

label fa_atr  ...  fa_mpenv_smooth total
weights 0.8  ...  0.5 NA
pose -1538.12  ...  -25.4072 -1234.64

In this case, you could pass 'pose' as the energy identifier, and SmoothT would take the last value '-1234.64', which represents the total score of the given PDB file. However, make sure that your PDB does not contain another line with the string 'pose' (after the correct line). Rosetta itself should not create another line containing 'pose'. Otherwise, rename the correct row to a unique identifier first.

If you use other software to calculate energies, make sure that the identifier is unique and that the correct value is at the end of that line. The identifier string does not necessarily have to be at the beginning of the line.

3. Upload / submit

Create a Zip archive from your ensemble of PDB files. This can have a size of up to 2 GB.

The Zip file has to contain PDB files of the starting and the final conformation. The names of these files have to entered into the form. The software will try to connect the start end end point with conformations from the ensemble.

The software first builds a graph. Nodes are the conformations. They will be connected by an edge if they are 'similar enough'. The mathematical criterion for similarity is the commonly used root mean squared deviation (RMSD). Please note, that the RMSD is calculated without superimposition.

Optionally, you can provide the chain. Only residues in that chain are considered for calculating the RMSD.

You have to provide the unique energy identifier to identify the lines in your PDB files containing the energy values.

4. Results

Once the pathway is calculated, you are forwarded to a result page. On the left the PDB files of the pathway are animated using the NGL viewer. On the right there is a plot of the energies associated with each conformation contained in the path.

The y-axis represents the energy values, and we denote the values of the x-axis as "relative RMSD". The relative RMSD is not really a continuous property in the strict sense. It is the sum of the RMSD values between successive conformations. A sphere indicates the conformation currently displayed in the 3D window. This moves synchronously with the 3D animation.

You can download both the resulting trajectory and the energy plot. There is also a link to the setting you used to create the animation. You can rotate and zoom the 3D view of the molecules, start and stop the animation. Additionally, you can open a full window view. Here, you have full access to all features of NGL. For further information about NGL see https://nglviewer.org