Sandra Romain

About me

PhD holder in bioinformatics. I completed my PhD in the GenScale team from the Inria Center of the University of Rennes (France), under the supervision of Claire Lemaitre (Inria/IRISA) and Fabrice Legeai (INRAE).

I am currently looking for a Post-Doctoral position.

• My CV: SandraRomain_CV_en.pdf
• You can contact me at: sandra [.] romain [.] pro [@] gmail [.] com

My PhD project

The heart of my project was to develop and implement methods for the identification and genotyping of genomic structural variants (SVs) using a pangenome representation. As a member of the Divalps ANR project team, I also performed data analysis on newly-produced long-read sequencing data from four species of Coenonympha butterfly, and worked towards the identification of large genomic rearrangements.

My PhD work can be divided into three main axes:

• Long-read SV genotyping using a variation graph representation (methodological development)
• Inversion characterization from pangenome graphs (benchmarking and methodological development)
• Inter-species genome comparison and inversion characterization (data analysis)

My full PhD thesis is available (in english) online: https://theses.hal.science/tel-04825910v1.

My PhD defense presentation slides: SandraRomain_PhD_Defense.pdf

Publications

• Sandra Romain, Thibaut Capblancq, Laurence Després, Mathieu Joron, Fabrice Legeai and Claire Lemaitre. (in prep.). Characterization of large inversions to investigate hybrid speciation in the four species-complex of alpine Coenonympha butterfly.
• Sandra Romain, Siegfried Dubois, Fabrice Legeai and Claire Lemaitre. (2025). Investigating the topological motifs of inversions in pangenome graphs. bioRxiv. doi.org/10.1101/2025.03.14.643331
• Fabrice Legeai, Sandra Romain, Thibaut Capblancq, Paul Doniol-Valcroze, Mathieu Joron, Claire Lemaitre and Laurence Després. (2024). Chromosome-level assembly and annotation of the pearly heath Coenonympha arcania butterfly genome. Genome Biology and Evolution, 16(3), evae055. doi.org/10.1093/gbe/evae055
• Sandra Romain and Claire Lemaitre. (2023). SVJedi-graph: improving the genotyping of close and overlapping structural variants with long reads using a variation graph. Bioinformatics, 39(Supplement_1), i270-i278. doi.org/10.1093/bioinformatics/btad237

Oral communications

• Sandra Romain and Claire Lemaitre. SVJedi-graph: improving the genotyping of close and overlapping structural variants with long reads using a variation graph. ISMB/ECCB (Intelligent Systems for Molecular Biology / European Conference on Computational Biology), July 2023, Lyon.
• Sandra Romain, Fabrice Legeai and Claire Lemaitre. Understanding the limits of pangenome graphs for the analysis of large inversions in a complex of butterfly species. EAGS (International Environmental and Agronomical Genomics Symposium), Feb. 2024, Toulouse.
• Sandra Romain and Claire Lemaitre. SVJedi-graph: using a variation graph to improve structural variant genotyping with long reads. DSB (Data Structures in Bioinformatics), June 2022, Düsseldorf.

Posters

• Sandra Romain and Claire Lemaitre. SVJedi-graph: genotyping close and overlapping structural variants with a variation graph and long reads. JOBIM (Journées Ouvertes en Biologie, Informatique et Mathématiques), July 2022, Rennes.

Softwares

SVJedi-graph

SVJedi-graph is the first variation-graph-based long-read structural variant (SV) genotyper. It is inspired from SVJedi (Lecompte et al., 2020), a long-read SV genotyper that overcomes reference bias during read mapping by representing both alleles of bi-allelic SVs as independant sequences. The use of a variation-graph representation of the SVs in SVJedi-graph improves both the genotyping rate and accuracy of close and overlapping SVs compared to state of the art tools.

• Implemented in Python
• Distributed under GNU Affero licence, Open Source
• Git: https://github.com/SandraLouise/SVJedi-graph

INVPG-annot

INVPG-annot is a tool that annotates inversion variants among the bubbles extracted from a pangenome graph.

• Implemented in Python
• Distributed under GNU Affero licence, Open Source
• Git: https://github.com/SandraLouise/INVPG_annot

Teaching experience

• Text algorithm and bioinformatics (University of Rennes, 2022)
  • First year of Computer Science Master
  • Lectures (6h), tutorial sessions (8h), practical sessions (10h) and student project supervision
• Algorithms, sequences and structures (University of Rennes, 2022)
  • Second year of Bioinformatics Master
  • Practical sessions (8h)

Other

• Sciences en Cour[t]s 2022 (festival of short films popularising science)
  • “Patatogène”, a 5-minute short film co-written and produced with fellow PhD students Baptise Ruiz, Kerian Thuillier, Khodor Hannoush and Roland Faure. Won the public prize of the festival.
  • Watch on youtube (in french): https://www.youtube.com/watch?v=9taXJ3P91YM&