In PLoS One this week, researchers at the Warsaw University of Life Sciences in Poland and elsewhere report a draft of the Cucumis sativus , better known as the cucumber, genome of the North-European Borszczagowski cultivar, line B10, as well as a comparative genomics analysis involving another cultivar, Arabidopsis thaliana, Populus trichocarpa, and Oryza sativa. Overall, the team found that "cucumber genomes show extensive chromosomal rearrangements, distinct differences in quantity of the particular genes … as well as in distributions of abscisic acid-, dehydration- and ethylene-responsive cis-regulatory elements in promoters of orthologous group of genes, which lead to the specific adaptation features." In particular, the team identified abscisic acid-specific cis-regulatory element distributions that suggest "why C. sativus is much more susceptible to moderate freezing stresses than A. thaliana," the authors write.
Over in PLoS Genetics, Johns Hopkins University's Hailiang Huang et al. present a "gene-wide significance test that uses greedy Bayesian model selection to identify the independent effects within a gene." The authors show that by applying their method to a data set of 2.5 million electrocardiography parameter-associated candidate SNPs in approximately 8,000 individuals, they were able to identify more validated associations than when using conventional genome-wide association study approaches. Further, "this method can be generalized to other study designs, retains power for low-frequency alleles, and provides gene-based p-values that are directly compatible for pathway-based meta-analysis," Huang et al. write.
Princeton University's Noorossadat Torabi and Leonid Kruglyak this week show that a coding polymorphism in TRM10 and a regulatory polymorphism in SUP45 affect translation termination efficiency in certain yeast strains. The duo used a quantitative dual luciferase reporter assay on two S. cerevisiae strains, finding that both "carry variants of TRM10 and SUP45 with opposite effects on translation termination efficiency." Further, Torabi and Kruglyak show that these "variants are common among 63 diverse S. cerevisiae strains and are in strong linkage disequilibrium with each other."
And in PLoS Computational Biology this week, a pair of researchers at Duke University present a Bayesian integration method to lower the false-positive and -negative misclassification rates associated with reconstructing protein-protein interaction networks, dubbed non-parametric Bayes ensemble learning, or NBEL. This approach automatically up-weights informative data sources, and down-weights less-informative and biased sources, as is "significantly more robust than the classic naïve Bayes to unreliable, error-prone and contaminated data," Duke's Chuanhua Xing and David Dunson write.