TY - JOUR T1 - The effect of statistical normalisation on network propagation scores JF - bioRxiv DO - 10.1101/2020.01.20.911842 SP - 2020.01.20.911842 AU - Sergio Picart-Armada AU - Wesley K. Thompson AU - Alfonso Buil AU - Alexandre Perera-Lluna Y1 - 2020/01/01 UR - http://biorxiv.org/content/early/2020/01/20/2020.01.20.911842.abstract N2 - Motivation Network diffusion and label propagation are fundamental tools in computational biology, with applications like gene-disease association, protein function prediction and module discovery. More recently, several publications have introduced a permutation analysis after the propagation process, due to concerns that network topology can bias diffusion scores. This opens the question of the statistical properties and the presence of bias of such diffusion processes in each of its applications. In this work, we characterised some common null models behind the permutation analysis and the statistical properties of the diffusion scores. We benchmarked seven diffusion scores on three case studies: synthetic signals on a yeast interactome, simulated differential gene expression on a protein-protein interaction network and prospective gene set prediction on another interaction network. For clarity, all the datasets were based on binary labels, but we also present theoretical results for quantitative labels.Results Diffusion scores starting from binary labels were affected by the label codification, and exhibited a problem-dependent topological bias that could be removed by the statistical normalisation. Parametric and non-parametric normalisation addressed both points by being codification-independent and by equalising the bias. We identified and quantified two sources of bias -mean value and variance- that yielded performance differences when normalising the scores. We provided closed formulae for both and showed how the null covariance is related to the spectral properties of the graph. Despite none of the proposed scores systematically outperformed the others, normalisation was preferred when the sought positive labels were not aligned with the bias. We conclude that the decision on bias removal should be problem and data-driven, i.e. based on a quantitative analysis of the bias and its relation to the positive entities.Availability The code is publicly available at https://github.com/b2slab/diffuBenchContact sergi.picart{at}upc.edu ER -