Viruses that infect animals can jump from one species to another, causing a new, usually severe disease in the new host. For example, in 2003 a virus in the Coronaviridae family jumped from an animal reservoir, believed to be horseshoe bats, to humans, causing a highly pathogenic disease in humans called severe acute respiratory syndrome (SARS). The ability of the SARS coronavirus to jump from horseshoe bats to humans undoubtedly required genetic changes in the virus. The changes are suspected to have occurred in the palm civet, since the SARS virus present in horseshoe bats is unable to infect humans directly.

Such newly emerged viruses often are highly infectious in humans, since they have not been previously encountered by the human immune system and thus humans have no immunity against them. The coronavirus that caused SARS, for example, spread quickly among humans, becoming a significant disease threat. The virus was quickly brought under control, owing to prohibitions on travel and quarantine measures. In late 2019 another type of coronavirus, called SARS-CoV-2, emerged in China and spread rapidly worldwide, giving rise to a pandemic. SARS-CoV-2 caused an illness known as coronavirus disease 2019 (COVID-19), which was very similar to SARS but caused significantly greater mortality, particularly among persons over age 65.

Influenza A viruses that infect humans can undergo a dramatic antigenic change, called antigenic shift, which generates viruses that cause pandemics. This dramatic change occurs because influenza A viruses have a large animal reservoir, wild aquatic birds. The RNA genome of influenza A viruses is in the form of eight segments. If an intermediate host, probably the pig, is simultaneously infected with a human and an avian influenza A virus, the genome RNA segments can be reassorted, yielding a new virus that has a surface protein that is immunologically distinct from that of influenza A viruses that have been circulating in the human population. Because the human population will have little or no immunological protection against the new virus, a pandemic will result. This is what most likely occurred in the 1957 flu pandemic, the 1968 flu pandemic, and the influenza pandemic (H1N1) of 2009.

Pandemic influenza A viruses can also apparently arise by a different mechanism. It has been postulated that the strain that caused the influenza epidemic of 1918–19 derived all eight RNA segments from an avian virus and that this virus then underwent multiple mutations in the process of adapting to mammalian cells. The bird flu viruses, which have spread from Asia to Europe and Africa since the 1990s, appear to be taking this route to pandemic capability. These viruses, which have been directly transmitted from chickens to humans, contain only avian genes and are highly pathogenic in humans, causing a mortality rate higher than 50 percent. Bird flu viruses have not yet acquired the ability to transmit efficiently from humans to humans, and it is not known what genetic changes must take place for them to do so.

Distinguishing taxonomic features

Viruses are classified on the basis of their nucleic acid content, their size, the shape of their protein capsid, and the presence of a surrounding lipoprotein envelope.

The primary taxonomic division is into two classes based on nucleic acid content: DNA viruses or RNA viruses. The DNA viruses are subdivided into those that contain either double-stranded or single-stranded DNA. The RNA viruses also are divided into those that contain double-stranded or single-stranded RNA. Further subdivision of the RNA viruses is based on whether the RNA genome is segmented. If the viruses contain single-stranded RNA as their genetic information, they are divided into positive-strand viruses if the RNA is of messenger sense (directly translatable into proteins) or negative-strand viruses if the RNA must be transcribed by a polymerase into mRNA.

All viruses falling into one of these nucleic acid classifications are further subdivided on the basis of whether the nucleocapsid (protein coat and enclosed nucleic acid) assumes a rodlike or a polygonal (usually icosahedral) shape. The icosahedral viruses are further subdivided into families on the basis of the number of capsomeres making up the capsids. Finally, all viruses fall into two classes depending on whether the nucleocapsid is surrounded by a lipoprotein envelope.

Some virologists adhere to a division of viruses into those that infect bacteria, plants, or animals; these classifications have some validity, particularly for the unique bacterial viruses with tails, but there is otherwise so much overlap that taxonomy based on hosts seems unworkable. Classification based on diseases caused by viruses also is not tenable, because closely related viruses frequently do not cause the same disease. Eventually, it is likely that the classification of viruses will be based on their nucleotide sequences and their mode of replication rather than on structural components, as is now the case.

The basic taxonomic group is called a family, designated by the suffix -viridae. The major taxonomic disagreement among virologists is whether to segregate viruses within a family into a specific genus and further subdivide them into species names. In the first decade of the 21st century, there occurred a shift toward the use of binomial nomenclature, dividing viruses into italicized genera and species. This move was prompted in large part by the International Committee on Taxonomy of Viruses (ICTV), a member group of the International Union of Microbiological Societies. The ICTV oversees the ongoing process of devising and maintaining a universal classification scheme for viruses. In the virus classification hierarchy, the ICTV recognizes orders, families, subfamilies, genera, and species. The placement of viruses in these groups is based on information provided by study groups composed of experts on specific types of viruses.

In the ICTV system, each species of virus is generally recognized as representing a group of isolates, or viruses with distinct nucleic acid sequences. Thus, a single species of virus may sometimes contain more than one isolate. Although the isolates of a species possess unique genetic sequences, they all descend from the same replicating lineage and therefore share particular genetic traits. Furthermore, isolates of a species also share in common the ability to thrive within a specific ecological niche. As scientists identify new isolates and species, the classification of viruses is expected to become increasingly complex. The following scheme presents examples of well-characterized DNA and RNA viruses as they are classified on the basis of the ICTV system.