A team from Taiwan and France presents findings from a genomic analysis of pre-domesticated Saccharomyces cerevisiae yeast lineages collected in Taiwan from the spring of 2016 to fall, 2020. Using whole-genome sequencing, the researchers assessed 121 S. cerevisiae isolates obtained by sampling hundreds of plants in more than a dozen forests with a range of climate, elevation, and plant types. The isolates fell into nine lineages, including three that appeared endemic to Taiwan, they report, providing a look at genetic diversity and selection patterns within isolates from the forests considered. "Collectively, Taiwanese isolates harbor genetic diversity comparable to that of the whole Asia continent, and different lineages have coexisted at a fine spatial scale even on the same tree," the authors report, noting that the yeast species "has rich natural diversity sheltered from human influences, making it a powerful model system in microbial ecology."
Memorial Sloan Kettering Cancer Center researchers describe GraphReg, a deep learning tool developed to model gene regulatory networks, including distal enhancers, based on measured chromatin interaction data. The method exploits 3D interactions from chromosome conformation capture assays to predict gene expression from one-dimensional epigenomic data or genomic DNA sequence, the team explains, noting that GraphReg appears to accurately predict gene expression with the help of graph attention networks that take enhancer element connections, epigenomic features, and DNA sequence features such as transcription factor binding sites into account. Moreover, the authors proposed still other iterations of the GraphReg approach, arguing that they "will have broad applicability for interpreting the function of epigenomic and genomic variation on gene expression."
A team from Stanford University, the Veteran Affairs Palo Alto Health Care System, and Washington University outlines an algorithm designed for digging into metagenomic microbial sequence data to characterize circular mobile elements. The approach, called DomCycle, "reconstructs likely circular genomes based on the identification of so-called 'dominant' graph cycles," the investigators write. After validating the strategy with simulated and real data, the authors applied DomCycle to 32 metagenomic sequence sets from environmental sampling sites, uncovering nearly 300 extrachromosomal circular mobile genetic elements that were subsequently clustered and analyzed more fully. "Clustering revealed 20 highly prevalent and cryptic plasmids that have clonal population structures with recent common ancestors," they write, adding that DomCycle "facilitates the study of microbial communities that evolve through horizontal gene transfer."