Inference of joint conformational distributions from separately-acquired experimental measurements

Abstract

Many biomolecules have flexible structures, requiring distributional estimates of their conformations. Experiments to acquire distributional data typically measure pairs of labels separately, losing information on the joint distribution. These data are assumed independent when estimating the conformational ensemble. We developed a method to estimate the true joint distribution from separately acquired measurements, testing it on two biological systems. This method accurately reproduces the joint distribution where known and generates testable predictions about complex conformational ensembles.