This repository contains code for predicting near-infrared (NIR) and UV-Vis properties of photoswitches using multitask learning with Chemprop.

Current prediction capabilties of photochemical properties

• Maximum absorption wavelength (nm)
• Extinction coefficient (log(M^-1 cm^-1))
• Photoisomerization quantum yield

If you use this software in your research, please cite it using the following:

@software{mt-nir,
  author = {Jesus Diaz Sanchez},
  title = {MT-NIR: Multi-task Photochemical Property Prediction Model},
  year = {2024},
  url = {https://github.com/jdsanc/mt-nir},
  version = {1.0.0}
}

You can also use the CITATION.cff file in the repository for citation information.

First, install Miniconda3 if you haven't already:

You can install Miniconda using Homebrew (recommended) or download directly:

# Install Homebrew if you don't have it
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"

# Install Miniconda
brew install --cask miniconda

# Download Miniconda
curl -O https://repo.anaconda.com/miniconda/Miniconda3-latest-MacOSX-x86_64.sh

# Install Miniconda
bash Miniconda3-latest-MacOSX-x86_64.sh

# Download Miniconda
wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh

# Install Miniconda
bash Miniconda3-latest-Linux-x86_64.sh

For Linux users, you might need to install wget first:

# Ubuntu/Debian
sudo apt-get install wget

# CentOS/RHEL
sudo yum install wget

# Fedora
sudo dnf install wget

Follow the prompts during installation. After installation, restart your terminal or run:

source ~/.bashrc  # Linux
source ~/.zshrc   # macOS

For more detailed Linux commands and troubleshooting, refer to the Linux Command Line Guide.

git clone https://github.com/jdsanc/mt-nir.git
cd mt-nir

You can create the environment in two ways:

conda env create -f environment.yml

conda create -n chemprop_v2 python=3.11
conda activate chemprop_v2
pip install chemprop pandas numpy rdkit

conda activate chemprop_v2

Model is already trained all you need to do load the model into your terminal. Using script described in Predictions.

You can handle both single SMILES strings and CSV files for bulk prediction:

For a single SMILES:

python predict.py --smiles "C1=CC=C(C=C1)N=NC2=CC=CC=C2"

Ensure you surround your input smiles by quotes ""

For a CSV file:

python predict.py --csv your_input_file.csv

Ensure if using your own .csv file to have the header written as "smiles" verbatum.

The script will output predictions in terminal for single prediction or in output csv called 'your_input_file_predict.csv' with the following properties:

• max_abs_wavelength (nm)
• extinct_coeff (log(M^-1 cm^-1))
• photoisomerization_QY