Profiling translated regions in viral genomes at scale

Abstract

Although viral genome sequencing has progressed rapidly, the functional annotation of elements such as translated regions is still incomplete for most viruses. Computational approaches face challenges in profiling open reading frames (ORFs) and experimental methods are limited by low throughput. Writing in Science, Weingarten-Gabbay et al. now report a massively parallel ribosome profiling method to screen for translated regions in hundreds of viruses in a single experiment. The authors transfected an oligonucleotide synthetic library — in which each oligonucleotide contained a 200-nucleotide viral sequence flanked by constant primers and cloned into an overexpression vector — into two human cell lines and then performed ribosome profiling. The oligonucleotides spanned the 5′ untranslated region and start of the coding sequence of 3,976 genes in 679 viral genomes.

Ribosome profiling identified 4,208 non-canonical ORFs. When the authors compared the pattern of ribosome footprints in the synthetic library with four native viral infections, they found a strong alignment between the location of footprints that originated from each analysis.