Section: New Results

Asymmetry tests for Bifurcating Auto-Regressive Processes with missing data

Participants : Benoîte de Saporta, Anne Gégout-Petit.

Using the properties of the estimators studied in [20] , we have constructed symmetry tests for bifurcating autoregressive processes (BAR) when some data are missing. BAR processes typically model cell division data. Each cell can be of one of two types odd or even. The goal of this work is to test asymmetry between odd and even cells in a single observed lineage. We have also derived asymmetry tests for the lineage itself, modeled by a two-type Galton-Watson process observed in a non standard scheme. We present applications on both simulated and real data.

This work is in collaboration with Laurence Marsalle of Lille 1 university. It will be soon submitted for publication.