Inferring relative surface elastic moduli in thin-wall models of single cells

TL;DR

This paper introduces a new inference scheme to map relative surface elastic moduli along single cell walls, improving spatial resolution and stability against noise through optimization methods and multiple sample analysis.

Contribution

It presents a novel primary inference scheme and two optimization schemes for more accurate and stable estimation of elastic moduli distributions in single cells.

Findings

The primary scheme is more stable with coarser marker spacing.

The first optimization scheme accurately captures canonical elastic modulus distributions.

The second scheme reliably predicts elastic modulus trends from a single cell.

Abstract

There is a growing interest in measuring the cell wall mechanical property at different locations in single walled cells. We present an inference scheme that maps relative surface elastic modulus distributions along the cell wall based on tracking the location of material marker points along the turgid and relaxed cell wall outline. A primary scheme provides a step-function inference of surface elastic moduli by computing the tensions and elastic stretches between material marker points. We perform stability analysis for the primary scheme against perturbations on the marker-point locations, which may occur due to image acquisition and processing from experiments. The perturbation analysis shows that the primary scheme is more stable to noise when the spacing between the marker points is coarser, and has been confirmed by the numerical experiments where we apply the primary scheme to synthetic cell outlines from simulations of hyper-elastic membrane deformation with random noise on the marker-point locations. To improve the spatial resolution of elastic modulus distribution of the primary scheme with noise, we propose two optimization schemes that convert the step-function inferences of elastic moduli into smooth-curve inferences. The first scheme infers a canonical elastic modulus distribution based on marker-point locations from multiple cell samples of the same cell type. The second scheme is a simplified cost-effective version that infers the elastic moduli based on marker-point locations from a single cell. The numerical experiments show that the first scheme significantly improves the inference precision for the underlying canonical elastic modulus distributions and can even capture some degree of nonlinearity when the underlying elastic modulus gradients are nonlinear. The second cost-effective scheme can predict the trend of the elastic modulus gradients consistently.