Installation and Dependencies

It is highly recommendable that TidyScreen is installed in an separate Conda virtual environment to prevent inconsistencies among the libraries currently installed in your system and those specifically required by the package itself:

# Create an isolated environment with the required libraries
$ conda create -n tidyscreen python=3.10 chemicalite

# Activate the corresponding environment
$ conda activate tidyscreen

# Install TidyScreen from the corresponding GitHub repository
$ pip install git+https://github.com/alfredoq/TidyScreen_v2

As part of compounds parameterization, TidyScreen implements a charge assignment method by taking profit of the accuracy and speed of a neural network-based approximation (Wang, Y; et.al.). As will be presented in the corresponding section, other charge system can be used, however it this so-called bcc-ml is highly recommended. In order to be able to use bcc-ml fitted charges, an additional library should be installed as follows in the TidyScreen environment:

# Install EspalomaCharge library
$ conda install -c conda-forge espaloma_charge openff-toolkit

In addition, in order to prepare molecules for docking studies TidyScreen workflows uses library called Meeko, which was developed by the ForliLab at Scripps Research Institute.Currently, TidyScreen requires a specific development version of Meeko capable of reading .mol2 atomic charges. This development version should be installed in the corresponding TidyScreen enviroment as follows:

# Install Meeko library
$ pip install git+https://github.com/forlilab/Meeko@develop