Skip to main content
Premium Trial:

Request an Annual Quote

This Week in Science: Sep 7, 2018

In this week's Science, a multi-institute team reports data showing minimal heterogeneity of genes driving metastases within an individual cancer patient, indicating that single biopsies can provide most of the information needed to inform treatment decisions. The researchers studied sequencing data for 76 untreated metastases from 20 patients, and inferred cancer phylogenies for breast, colorectal, endometrial, gastric, lung, melanoma, pancreatic, and prostate cancers. Within individual patients, they find that "a large majority of driver gene mutations are common to all metastases." Notably,  the driver gene mutations that were not shared by all metastases were unlikely to have functional consequences. The researchers also present a mathematical model of tumor evolution and metastasis formation to explain their observations. GenomeWeb has more on this, here.

Also in Science, two separate research groups publish studies identifying proteins that inhibit the activity of Cas12a, an enzyme-guided RNA that has proven to be a highly effective alternative to Cas9 in CRISPR genome editing. In the first paper, the University of California, Berkeley’s Jennifer Doudna and colleagues use a bioinformatic and experimental screening approach to identify three different inhibitors that block or diminish CRISPR-Cas12a-mediated gene editing in human cells. The methodology, they say, could likely be used to find similar anti-CRISPR proteins in other organisms. In the second paper, a University of California, San Francisco-led team uses bioinformatics to screen for anti-CRISPR proteins and also find a number of them that block CRISPR-Cas12a genome editing. GenomeWeb also has more on these papers, here.

And in Science Translational Medicine, investigators from Personal Genome Diagnostics and elsewhere report on a machine-learning approach for discovering somatic mutations in tumors. Called Cerebro, the method is designed to identify such mutations in subclonal or low-purity samples, as well as differentiate them from germline changes or PCR and next-generation sequencing artifacts. The researchers find that Cerebro outperforms existing methods in sensitivity and predictive value in both simulated and experimentally validated cancer genomes. They also demonstrate the tool's utility using paired tumor-normal exome data from 1,368 patients included in The Cancer Genome Atlas, and demonstrate how its integration into a clinical sequencing platform can boost mutation detection. "Our approach, which is entirely automated, would eliminate the need for expert review of sequence data, a practice commonly used in the clinical setting and likely unsustainable for widespread NGS analyses," the authors write.