This repository contains the code for the discovery of hyperspectral reporter molecules and the processing of collected hyperspectral images. The code is organized into directories that represent standalone modules for the different aspects of the work. Each directory contains a README file to guide users.
01_biochemical_network: Scripts and notebooks to process biomolecular reaction databases, extract metabolites, define host metabolomes, construct biochemical reaction network and perform shortest path search to assess ease of biosynthesis of heterolous metabolites02_spectral_prediction: Scripts to run TD-DFT spectral prediction pipeline to go from metabolite structures to predicted spectra.03_reporter_ranking: Scripts to compute the uniqueness of molecules compared to other metabolites and to compute contrast to hyperspectral images. Also contains benchmarking scripts.04_image_processing: Scripts to classify presence of reporter in experimentally collected hyperspectral images.