Inverse Design of Singlet Fission Materials with Uncertainty-Controlled Genetic Optimization

Overview

This repository contains the implementation of the uncertainty-controlled genetic algorithm (ucGA) for the design of singlet fission materials presented in the paper (TODO:Link).

Installation

1. Clone the repository:

   git clone https://github.com/lcmd-epfl/ucGA.git
   cd ucGA

2. Create and activate the Conda environment:

   conda env create -f env.yml
   conda activate ucGA

3. Further Installations: Install NaviCatGA, theodore and QML by following the instructions on the respective project websites.

4. SCScore: Install the SCScore to utils/.

Usage

1. Configure the parameters: Edit config.py to set the appropriate paths for input data, models, and output directories.

2. Run the main optimization:

   python main.py <output_folder>

Important Files

• config.py: Configuration file for setting paths and parameters for the genetic optimization. Edit this file to specify paths for input data, models, and output directories according to your system's directory structure.
• main.py: Main script for launching the uncertainty-aware optimization.

Repository Structure

• assembler/: Automatically assembling molecules from the reFORMED database.
• fitness_evaluation/: Uncertainty-aware fitness function determination.
• model_predictors/: Prediction of properties using SLATM models.
• objective_scores/: Calculate scores to assess singlet fission propensity.
• quantum_calculations/: Scripts for performing xTB and TD-DFT calculations.
• utils/: Utility scripts and functions.

Citation

If you use this code, please cite the following paper:

Luca Schaufelberger, TODO

Contact

For questions, please contact `schaluca@ethz.ch'