Problems with the tutoria

[caoyx061]

After installing the latest version of OpenFE, I encountered two problems while running the notebook tutorial. The first error message was "ValueError: Program neato not found in path.", and the second indicated that the ambertools toolkit could not be used. These softwares were installed correctly during the OpenFE installation process. What could be the cause of these issues?

[IAlibay]

@caoyx061 thanks for reporting this, could I ask for a little bit more information about the issues you saw?

1. Which notebook did you run?
2. How did you install openfe? If via mamba could you provide the output of mamba list here?
3. Could you post the full traceback for both the errors you saw?

[caoyx061]

I am thrilled to receive your answer. In response to your questions, here are my detailed replies:

While using the notebook for the "Setting up a relative binding free energy network" project, I encountered an issue when calling plot_atommapping_network(ligand_network), which resulted in a ValueError: "Program neato not found in path."
I installed OpenFE following the tutorial, using mamba activate openfe_env.
Since the issue has been resolved, I can't provide relevant logs here.
To fix the problem, I discovered that the neato file was empty after installation. After searching online, I found others had encountered the same issue. After trying several methods, I ultimately resolved it by running the following commands in the conda environment:

sudo apt-get install graphviz graphviz-dev
pip install pygraphviz

​
sudo apt-get install graphviz graphviz-dev
pip install pygraphviz
This installed the complete neato file, and I was able to display it correctly.

I also have two small questions I’d like to ask you:

When running RBFE, only one CPU core is being utilized while the others remain idle, and GPU usage is at 90%. I run a complex task that takes about 4 hours. My GPU is a 409, and my CPU model is a 12 vCPU Intel(R) Xeon(R) Platinum 8352V. Is this normal?
The documentation on preparing ligand files and constructing a reasonable network seems lacking. Did I overlook something? I would greatly appreciate any related tutorials.

I sincerely hope that your software continues to improve.​