The challenge in detecting a set of viruses simultaneously in clinical samples lies in the large variety of their genomes, i.e. there is no gene that is present in all viruses. With the advent of new sequencing technologies, including high throughput parallel sequence analysis, it became possible to process more sequence data. So much more in fact that now there might be a new way to detect multiple viruses at once in one clinical sample, as Wylie et al report.
The report documents a computational screening of about 190 viral genera and 337 viral species, excluding human endogenous retroviruses, which reduced the number of base pairs to less than 200 million of reference sequences. After k-mer and cluster analysis, a unique set of sequences emerged.
Subsequently, targeted sequence capture, a technique by which viral sequences in a particular sample are enriched with specific nucleic acids, was applied. This technique is often used to overcome some of the limitations of metagenomic shotgun sequencing. One of those limitations includes the amplification of sequences of viruses present in low numbers/titers in e.g. asymptomatic or healthy individuals.
The resulting reference panel was thus tested in the clinic and the authors conclude that this technique has the potential to be used as a diagnostic tool to detect viruses in clinical samples.
The advantage of this technique over microarrays is that it provides sequence information of the virus in question.
The full version of the article can be found on:
http://genome.cshlp.org/content/early/2015/09/22/gr.191049.115.full.pdf+html