The identification of new virus species is a key issue for the study of infectious disease but is technically very difficult. We developed a system for large-scale molecular virus screening of clinical samples based on host DNA depletion, random PCR amplification, large-scale sequencing, and bioinformatics. The technology was applied to pooled human respiratory tract samples. The first experiments detected seven human virus species without the use of any specific reagent. Among the detected viruses were one coronavirus and one parvovirus, both of which were at that time uncharacterized. The parvovirus, provisionally named human bocavirus, was in a retrospective clinical study detected in 17 additional patients and associated with lower respiratory tract infections in children. The molecular virus screening procedure provides a general culture-independent solution to the problem of detecting unknown virus species in single or pooled samples. We suggest that a systematic exploration of the viruses that infect humans, ''the human virome,'' can be initiated.bioinformatics ͉ nucleotide sequencing ͉ respiratory tract infection ͉ virus V irus infections impose an enormous disease burden on humanity, but our knowledge of the viruses that infect humans is still incomplete. Because of the elusive nature of viruses, most studies are limited to the investigation of already known viruses, whereas the discovery of an unknown virus and production of the first diagnostic reagent is very difficult and remains a rare occurrence. The majority of viruses known today were first identified by animal experiments or virus replication in tissue culture. We know that there are viruses that cannot be replicated in the laboratory, and some of these have been identified by molecular methods (1-4). No generally applicable method for the identification of such viruses has been available, and, in fact, a very large number of unidentified human viruses may exist (2). Thus, numerous acute and chronic diseases with unknown etiology may be caused by unidentified viruses, and the systematic search for unknown viruses is an urgent scientific task (2).Lower respiratory tract infection (LRTI) is a leading cause for hospitalization of infants and young children and accounts for Ϸ250,000 hospitalizations a year in the United States alone (5). The most important viral agent in this group of patients is respiratory syncytial virus (RSV). Other important agents are influenza viruses, parainfluenza viruses, adenoviruses, rhinoviruses, coronaviruses, and human metapneumovirus (6-8). Human metapneumovirus was cultured and characterized in 2001, and, more recently, a renewed interest in coronaviruses after the severe acute respiratory syndrome epidemic has led to the characterization of two coronavirus species associated with LRTI (9-11). In comprehensive studies of the etiology of LRTI, no etiologic agent has been found in 12-39% of cases, which suggests that additional unknown agents may be involved in the etiology of LRTI (6-8).The identification of viruses...

Human bocavirus is prevalent among children with acute wheezing and can cause systemic infection. Results suggest a model for bocavirus infection in which high viral loads are potentially associated with respiratory symptoms and low viral loads indicate asymptomatic shedding. Therefore, quantitative polymerase chain reaction analysis may be important for additional studies of human bocavirus.

By supplementing traditional diagnostic methods with new PCR-based methods, a high microbial yield was achieved. This was especially evident for patients with complete sampling. Mixed infections were frequent (most commonly S. pneumoniae together with a respiratory virus).