A modeling study of the impact of treatment policies on the evolution of resistance in sea lice on salmon farms.

Salmonid aquaculture is an important source of nutritious food with more than 2 million tonnes of fish produced each year (Food and Agriculture Organisation of the United Nations, 2019). In most salmon producing countries, sea lice represent a major barrier to the sustainability of salmonid aquaculture. This issue is exacerbated by widespread resistance to chemical treatments on both sides of the Atlantic. Regulation for sea lice management mostly involves reporting lice counts and treatment thresholds, which depending on interpretation may encourage preemptive treatments. We have developed a stochastic simulation model of sea lice infestation including the lice life-cycle, genetic resistance to treatment, a wildlife reservoir, salmon growth and stocking practices in the context of infestation, and coordination of treatment between farms. Farms report infestation levels to a central organisation, and may then cooperate or not when coordinated treatment is triggered. Treatment practice then impacts the level of resistance in the surrounding sea lice population. Our simulation finds that treatment drives selection for resistance and coordination between managers is key. We also find that position in the hydrologically-derived network of farms can impact individual farm infestation levels and the topology of this network can impact overall infestation and resistance. We show how coordination and triggering of treatment alongside varying hydrological topology of farm connections affects the evolution of lice resistance, and thus optimise salmon quality within socio-economic and environmental constraints. Network topology drives infestation levels in cages, treatments, and hence treatment-driven resistance. Thus farmer behaviour may be highly dependent on hydrologically position and local level of infestation.

Genome-wide aggregated trans-effects on risk of type 1 diabetes: A test of the "omnigenic" sparse effector hypothesis of complex trait genetics.

The "omnigenic" hypothesis postulates that the polygenic effects of common SNPs on a typical complex trait are mediated through trans-effects on expression of a relatively sparse set of effector ("core") genes. We tested this hypothesis in a study of 4,964 cases of type 1 diabetes (T1D) and 7,497 controls by using summary statistics to calculate aggregated (excluding the HLA region) trans-scores for gene expression in blood. From associations of T1D with aggregated trans-scores, nine putative core genes were identified, of which three-STAT1, CTLA4 and FOXP3-are genes in which variants cause monogenic forms of autoimmune diabetes. Seven of these genes affect the activity of regulatory T cells, and two are involved in immune responses to microbial lipids. Four T1D-associated genomic regions could be identified as master regulators via trans-effects on gene expression. These results support the sparse effector hypothesis and reshape our understanding of the genetic architecture of T1D.