Layer-by-layer decoding of contemporary and historic painting structure and composition using MALDI mass spectrometry imaging and machine learning

This repository contains the MSIpredictART model used in the above publication: Layer-by-layer decoding of contemporary and historic painting structure and composition using MALDI mass spectrometry imaging and machine learning. MSIpredictART is a predictive model for identification of binder and pigment composition of paint cross-sections based on Least Absolute Schrinkage and selection Operator (LASSO). MSIpredictART can accurately identify with >95% specificity range of commonly occurring historical inorganic pigments and organic binders both oil and protein based.