Pairwise Difference Regression: A Machine Learning Meta-algorithm for Improved Prediction and Uncertainty Quantification in Chemical Search

Abstract

Machine learning (ML) plays a growing role in the design and discovery of chemicals, aiming to reduce the need to perform expensive experiments and simulations. ML for such applications is promising but difficult, as models must generalize to vast chemical spaces from small training sets and must have reliable uncertainty quantification metrics to identify and prioritize unexplored regions. Ab initio computational chemistry and chemical intuition alike often take advantage of differences between chemical conditions, rather than their absolute structure or state, to generate more reliable results. We have developed an analogous comparison-based approach for ML regression, called pairwise difference regression (PADRE), which is applicable to arbitrary underlying learning models and operates on pairs of input data points. During training, the model learns to predict differences between all possible pairs of input points. During prediction, the test points are paired with all training set points, giving rise to a set of predictions that can be treated as a distribution of which the mean is treated as a final prediction and the dispersion is treated as an uncertainty measure. Pairwise difference regression was shown to reliably improve the performance of the random forest algorithm across five chemical ML tasks. Additionally, the pair-derived dispersion is both well correlated with model error and performs well in active learning. We also show that this method is competitive with state-of-the-art neural network techniques. Thus, pairwise difference regression is a promising tool for candidate selection algorithms used in chemical discovery.

Uncertainty quantification: Can we trust artificial intelligence in drug discovery?

“…Tynes et al. transferred this idea to molecular property prediction and observed similar results ( Tynes et al., 2021 ).…”

“…By conducting this prediction procedure for all reference ligands and the new ligand, multiple predicted values of its pIC 50 could be obtained and the variance of these predicted values could be regarded as an estimate of the uncertainty. Tynes et al transferred this idea to molecular property prediction and observed similar results (Tynes et al, 2021).…”

DeepDelta: Predicting Pharmacokinetic Improvements of Molecular Derivatives with Deep Learning

“…Several other molecular pairing approaches have been deployed for various purposes. For example, the pairwise difference regression (PADRE) approach trains machine learning models on pairs of feature vectors to improve the predictions of absolute property values and their uncertainty estimation 22 . Similarly, AstraZeneca has created workflows that utilize compound pairs to train Siamese neural networks for compound classification and regression.…”

Physicochemical Responsive Integrated Similarity Measure (PRISM) for a Comprehensive Quantitative Perspective of Sample Similarity Dynamically Assessed with NIR Spectra

“…Comparing each source sample allows PRISM to maximize capturing similarity information embedded in the source domain covariance structure relative to target domain samples. Differences have been used in other studies to expand and better detail the sought information for other purposes. ,− …”