A new bioinformatics analysis method developed at the Department of Virology of Julius-Maximilians-Universität Würzburg (JMU) in Germany has provided us with new information regarding viral infections. The method, published in the Nature Methods, used to demonstrate that virus-infected cells produce far more infection-related proteins and peptides than previously thought.


An infection with cytomegalovirus is usually harmless for adults. However, during pregnancy the virus can be transmitted to the unborn baby and cause malformations. Once the viruses have invaded a human cell, they start to produce large amounts of viral proteins. This includes more than 500 different proteins and peptides, including 200 previously unknown to science.

The new technique allows the ribosomal activities to be captured much more accurately than before. All proteins and peptides are assembled at the numerous ribosomes of a cell. During viral infection, the ribosomes also synthesize all proteins the virus needs to reproduce. The assembly instructions are delivered by special messenger molecules, the mRNAs.

Which proteins and peptides are produced at the ribosomes of a cell and in which amounts? How does this profile change due to the stress of viral infection? These questions can be answered by state-of-the-art high-throughput sequencing methods that use ribosomal profiling (Ribo-seq). They visualize ribosomal activities, the so-called translation processes, as periodic patterns.

"Previously, a number of error sources often prevented the reliable detection of translation events when analyzing Ribo-seq data," says Florian Erhard. What is more, in at least half of all mRNAs, shorter open reading frames (sORFs) precede the known open reading frames (ORFs). Very difficult to pinpoint in the measured data, they are however an important cellular regulating mechanism, especially under stress.

The new bioinformatics procedure called PRICE substantially improves the identification of translation events. "Our method is capable of resolving even complex cases, for example overlapping ORFs or unusual start codons, with high accuracy. This has allowed us to determine all translated areas genome-wide with high accuracy for the first time," Professor Erhard explains.

As a result, the team discovered various new cellular and viral peptides. Moreover, the researchers observed that hundreds of sORF peptides are efficiently presented at the cell surface by MHC-I molecules. "sORFs thus encode for a new class of antigens that can be recognized by our immune system," says Lars Dölken. "We, therefore, assume that sORFs are involved in immunological control mechanisms especially during virus infections and stress responses." 

"PRICE now enables us to analyze all existing and future dataset in much greater depth and with substantially improved accuracy," Dölken says further. He believes that the gain is so big that it justifies re-analyzing all previously published data. The researchers assume that their method will be used on a wide scale and become the international standard to analyze Ribo-seq experiments.

"We are convinced that the enhanced data analysis will deliver essential new insights in many fields of biomedical research," the researcher concluded. In virology, the new method may contribute to getting a better understanding of cytomegalovirus. The new technique is relevant for medical applications, because knowledge of the repertoire of viral proteins used to evade the immune system, for example, is crucial to fight infections or develop vaccines.