Ajithkumar Munusamy ^1* Jonathan D’Ambrosio ² Marie-Agnès Travers ³ Romain Morvezen ² Lionel Dégremont ¹

• ¹ IFREMER, La Tremblade, France
• ² SYSAAF, Rennes, France
• ³ UMR5244 Interactions Hôtes Pathogènes Environnements (IHPE), Perpignan, Languedoc-Roussillon, France

The blue mussel is one of the major aquaculture species worldwide. In France, this species faces a significant threat from infectious disease outbreaks in both mussel farms and the natural environment over the past decade. Diseases caused by various pathogens, particularly Vibrio spp., have posed a significant challenge to the mussel industry. Genetic improvement of disease resistance can be an effective approach to overcoming this issue. In this work, we tested genomic selection in the blue mussel (Mytilus edulis) to understand the genetic basis of resistance to one pathogenic strain of Vibrio splendidus (strain 14/053 2T1) and to predict the accuracy of selection using both pedigree and genomic information. Additionally, we performed a genome-wide association study (GWAS) to identify putative QTLs underlying disease resistance. We conducted an experimental infection involving 2,280 mussels sampled from 24 half-sib families containing each two full-sib families which were injected with V. splendidus. Dead and survivor mussels were all sampled, and among them, 348 dead and 348 surviving mussels were genotyped using a recently published multi-species medium-density 60K SNP array. From potentially 23.5K SNPs for M. edulis present on the array, we identified 3,406 high-quality SNPs, out of which 2,204 SNPs were successfully mapped onto the recently published reference genome. Heritability for resistance to V. splendidus was moderate ranging from 0.22 to 0.31 for a pedigree-based model and from 0.28 to 0.36 for a genomic-based model. GWAS revealed the polygenic architecture of the resistance trait in the blue mussel. The genomic selection models studied showed overall better performance than the pedigree-based model in terms of accuracy of breeding values prediction. This work provides insights into the genetic basis of resistance to V. splendidus and exemplifies the potential of genomic selection in family-based breeding programs in M. edulis.