A University of Copenhagen-lead team considers genome architecture and regulatory features shaping somatic structural variant patterns. Applying Hi-C and other analyses to a human cell line, the researchers characterized chromatin interactions, related SV events, and the gene regulation effects of both processes in cells that had undergone widespread and complex rearrangements through chromothripsis. Along with interactions identified there, they analyzed more than 2,700 patient-derived cancer genomes to further tease out three-dimensional chromatin biases impacting the formation of SVs. "We show that SVs have a strong tendency to occur between similar chromatin compartments and replication timing regions," the authors report, noting that "SVs frequently occur at 3D loop-anchors, that SVs can cause a switch in chromatin compartments and replication timing, and that this is a major source of SV-mediated effects on nearby gene expression changes."
German researchers describe an analytical framework called CADD-SV for bringing together structural variant annotation clues to predict the phenotypic effects of SVs, including potential ties to health or disease. At the center of the approach, the "Combined Annotation Dependent Depletion" (CADD) machine learning method makes it possible to assess SV effects based on available annotation data, the team says, prioritizing potentially deleterious SVs with summary statistic computation and random forest modeling. "To validate this new approach, we apply CADD-SV to distinguish common SVs from annotated disease-causing variants and to identify functional variants on independent datasets of germline and somatic SV," the author write. "Our tool can be used to highlight disease-causing SVs in supposedly healthy individuals … and allows [us] to prioritize regulatory, non-coding variants like expression Quantitative Trait Loci (eQTLs) or variants under natural selection."
Finally, researchers in Australia, the US, France, and the UK present findings from a metabolic analysis of taxa in the Klebsiella pneumoniae species complex (KpSC), known for causing opportunistic healthcare infections. "Due to increasing rates of multi-drug resistance within the KpSC, there is a growing interest in better understanding the biology and metabolism of these organisms to inform novel control strategies," they note. The team came up with strain-specific genome-scale metabolic models (GEMs) based on genome sequences for 37 KpSC isolates spanning all seven KpSC taxa, incorporating substrate-specific growth phenotype simulations and real phenotypic data to flag essential genes and substrate use patterns in the strains. "These analyses revealed multiple strain-specific differences, within and between species, and highlight the importance of selecting a diverse range of strains when exploring KpSC metabolism," the authors report, adding that their GEM set "could be used to inform novel drug design, enhance genomic analyses, and identify novel virulence and resistance determinants."