While exploring mechanisms behind the CpG island methylator phenotype (CIMP) in gastric cancer, researchers from Duke-NUS Medical School, the National University of Singapore, and elsewhere describe related recurrent epigenetic mutations, or "epimutations," targeting the beta-synthase (CBS) enzyme. The team started by doing methylation sequencing, MeDIP, RNA sequencing, and mass spec-based proteomic profiling in more than a dozen gastric cancer cell lines, uncovering CIMP-related epigenetic silencing of CBS and other features that were subsequently assessed with similar analyses on 50 additional gastric cancer cell lines and 467 primary gastric cancer samples. Some of the same DNA methylation shifts turned up when the authors removed the CBS-coding gene from normal gastric epithelial cells, while mouse models with lower-than-usual CBS levels showed immune-related expression changes in stomach tissue. "Reflecting its metabolic role as a gatekeeper interlinking the methionine and homocysteine cycles, CBS loss in vitro causes reductions in the anti-inflammatory gasotransmitter hydrogen sulfide (H2S), with concomitant increase in NF-[kappa B] activity," they report, noting that the results point to "H2S donors as a potential new therapy for CBS-silenced lesions."
A team from Sweden, Denmark, and the US describes a correction method aimed at better quantifying molecules in single-cell RNA sequencing assays that rely on unique molecular identifiers (UMIs). The BUTTERFLY approach is intended to address the so-called "pooled amplification paradox," the researchers say, a situation stemming from incomplete correction for amplification biases that can occur after simply removing duplicate UMIs. In contrast, they say, analyses on several specific genes and cell types suggest that the BUTTERFLY correction appears to provide a more precise look at molecules assessed by single-cell RNA-seq. "In addition to improving abundance estimates of specific genes, we have shown that BUTTERFLY can help reduce batch effects between datasets sequenced at different depths," they write, noting that the broader approach behind BUTTERFLY "is relevant for any assay in which objects are sampled after amplification, and where the pooled amplification paradox may occur."
Finally, investigators in Germany, China, Switzerland, and South Korea outline their "Genome UNClutterer," or GUNC, workflow for finding chimeric genome sequences and removing sequence contaminants in prokaryotic genome sets or metagenomic data — an approach the team applied to simulated data and microbial sequences in large databases such as GenBank. "GUNC complements existing approaches by targeting previously under-detected types of contamination," the authors write, adding that "[w]e expect that an automated, rapid, and accurate quantification of genome contamination will further enable genome-centric microbiology at large scale and high resolution."