Skip to main content
Premium Trial:

Request an Annual Quote

This Week in PNAS: Jul 10, 2018

In the early, online edition of the Proceedings of the National Academy of Sciences, researchers from the US and China describe a computational approach for bringing together and integrating different types of genomic data from individuals cells, including single-cell RNA sequences and single-cell ATAC-seq data. The method — known as coupled non-negative matrix factorizations, or coupled NMF — builds on a so-called 'coupled clustering' problem formulated as an optimization problem, the team says, noting that they "develop[ed] an approach for its solution based on the coupling of two non-negative matrix factorizations." After applying the NMF clustering approach to simulated single-cell data, the team applied this strategy to integrate single-cell RNA-seq and single-cell ATAC-seq data from mouse embryonic stem cells treated with retinoic acid.

A team from the University of Notre Dame and Indiana University present findings from a systems genetic analysis focused on characterizing the consequences of two chromosome 2 inversion polymorphisms — believed to represent local adaptations — in the malaria-carrying African mosquito Anopheles gambiae. When they considered transcriptional profiles, behavior, and physiological traits in An. gambiae mosquitos from four karyotypes, including mosquitos exposed to rapamycin and raised in different environments, the researchers found that "[a]cclimation to different climatic regimes resulted in pervasive inversion-driven phenotypic differences whose magnitude and direction depended upon gender, environment, and epistatic interactions between inversions."

Researchers from the Hebrew University of Jerusalem and other centers report on a quantitative method for developing patient-specific transcriptional networks from large-scale cancer genome data. The strategy "identifies signatures comprising patient-specific oncogenic processes rather than cancer type-specific biomarkers," the authors write, adding that "comprehensive transcriptional signatures should allow for more accurate classification of cancer patient and better patient-specific diagnostics." When the team applied the network approach to 527 lymphoma, bladder cancer, gastric cancer, colorectal cancer, and breast cancer samples, together with normal stomach tissue, it highlighted key molecular processes in the tumors as well as alteration combinations that appeared to differentiate individual patient transcriptional networks.