Remarks on the statistical study of protein-protein interaction in living cells

TL;DR

This paper compares mono- and bi-exponential models for protein-protein interactions in living cells, revisiting classical statistical methods to enhance analysis with small biological datasets and discussing inherent limitations.

Contribution

It introduces improved statistical approaches for model selection in small-sample biological data and highlights the limitations of current methods.

Findings

Bi-exponential models better fit certain protein interaction data.

Classical statistical methods can be refined for small biological datasets.

Limitations of current models are identified and discussed.

Abstract

In this note, we focus on a selection model problem: a mono-exponential model versus a bi-exponential one. This is done in the biological context of living cells, where small data are available. Classical statistics are revisited to improve existing results. Some unavoidable limits are also pointed out.