Viral infections impose an enormous disease burden on humanity, but our knowledge about pathogenic viruses is rather incomplete. Even well known viruses such as influenza may surprise us. The recent outbreak of the novel pandemic strain A(H1N1), also known as "swine flu", has been a perfect example. Almost 50,000 people around the globe have been infected within a relatively short time, and on 11th June 2009 the World Health Organization declared a global flu pandemic.

In recent years several other novel, life-threatening bugs have emerged-HIV, Ebola, highly pathogenic avian influenza and SARS are just few examples. You could ask, "who would want to study something as ordinary as the common cold?" On the other hand, how many times have you actually suffered from Ebola compared with the number of sneezing and coughing periods you've had in the past year?

Actually, the story to tell is not about sneezing itself, but about the hunt for pathogens that make us sneeze. Despite the indisputable progress of medical science, we often simply have no idea what is causing a patient's disease. When it comes to respiratory infections, this point might apply to as many as 30% of all patients.

In order to break this status quo, research centres are continuously searching for new pathogens in material from patients with atypical illnesses or infections that aren't straightforward to diagnose. Several methods have been developed to hunt down and identify novel pathogens. There are considerable problems with current techniques though-we do not know what we are looking for, we do not know what it looks like and we even do not know if it is there!

Together with Dr Lia van der Hoek and Professor Ben Berkhout, my research has made an important contribution to this field. After many long evenings and several failures, we developed a new technique we called ‘VIDISCA' (an abbreviation for ‘virus discovery based on cDNA-AFLP'). The method is fairly simple-we copy all kinds of genetic material into double stranded DNA and subsequently chop it up with enzymes called restriction enzymes. How a certain material will be cut depends on its properties; therefore, every virus will provide us with a unique pattern of different size pieces. Next, we glue in synthetic parts at the cut sites to flank the unknown material. These artificial regions are further used to amplify and interpret the material, allowing us to instantly recognise whether we have managed to find anything interesting.

Shortly after the SARS-CoV outbreak, we used this method to identify a novel human coronavirus we named NL63, the fourth in the coronavirus family. This virus belongs to the same group as the SARS virus and causes respiratory illness in infected patients. Further studies have shown that NL63 is also the major cause of croup-a respiratory syndrome characterised by a dry, barking cough that worsens at night, known to all mothers with young children.

How important is this discovery? Well, NL63 is a human pathogen that infects virtually everybody during his or her lifetime. And if that is not enough, let's put the finding in context. It took several months before the SARS coronavirus was recognised, as diagnostic methods at the time were not sophisticated enough to detect the virus. Our knowledge of the human coronavirus NL63 and another one discovered in 2005-HKU1-provided us with crucial information about variation among coronaviruses that has been used to facilitate the design of new techniques and models for detection of other more pathogenic variants.

On the other hand, identification of a novel pathogen is just the beginning of a new story. How does the virus replicate? How does it interact with the infected host and with other pathogens? Where does it come from? What disease does it cause? And, ultimately, how can we stop it? We have set up several molecular biology projects to delve into all these questions, but reaching the answers will take years of careful studies.

And for now? All available methods of identifying pathogens are either very accurate or able to identify a broad spectrum of material. Never both. To overcome this limitation, we are trying to combine computer programming with more advanced ways of identifying pathogens. By limiting the restrictions on what is labelled as interesting, we can increase the variety of pathogens that we may identify. By limiting points where the process might fail, we improve the accuracy of identification. And, step by step, we will learn how to carefully and efficiently control bugs in the future. (Krzysztof Pyrć, Jagiellonian University in Krakow,