Aller au contenu. | Aller à la navigation

Outils personnels

Navigation
Vous êtes ici : Accueil / Actualités / Séminaires / Séminaires internes / Nicolas Mendiboure

Nicolas Mendiboure

Computational frameworks to decipher homology search during DNA break repair by homologous recombination
Quand ? Le 10/07/2026,
de 11:00 à 12:00
S'adresser à Equipe Piazza, Salle des thèses
Ajouter un événement au calendrier vCal
iCal

Homologous recombination (HR) is a conserved DNA repair pathway that uses an intact homologous molecule as a template to mend double-strand breaks and single-stranded gaps. Its accuracy depends on how the broken DNA searches for and selects a homologous donor in the nuclear space. Choosing a wrong, non-allelic donor can generate structural variants, so the homology search must be both efficient and faithful. The mechanisms that ensure this balance remain incompletely understood.

 

My PhD combines data analysis and modeling to dissect these mechanisms in S. cerevisiae. First, I developed sshicstuff, a pipeline to analyse ssHi-C data, a Hi-C variant that captures the single-stranded DNA contacts made by the Rad51 filament during the search, which standard Hi-C cannot see. Second, I built a stochastic model of Rad51 filament assembly and disassembly, showing how the interplay between Rad55-Rad57 and Srs2 drives the growth of long, stiff filaments competent for genome-wide search. Third, I developed SHERPA, a spatial model that integrates filament kinetics, protein binding on ssDNA, and the nuclear geometry inferred from ssHi-C. It simulates thousands of virtual cells and outputs first passage times, D-loop levels, and contact maps directly comparable to experiment.