Australasian Science: Australia's authority on science since 1938

Predicting Pandemics

Credit: Gino Santa Maria

Credit: Gino Santa Maria

By Jemma Geoghegan

Which factors determine whether an emerging virus is likely to burn out or spread like wildfire between people?

The vast majority of “emerging” viruses capable of infecting humans have “jumped” from animals to humans. This includes some of the most devastating epidemics on record, such as the ongoing global HIV/AIDS pandemic.

Fortunately it seems that most new viruses that jump from animals to humans are incapable of spreading among the human population. For example, bird flu has repeatedly jumped from poultry to humans, but has not yet adapted to spread directly from those humans to others.

However, some viruses have been able to establish human-to-human transmission after an initial jump from animals, including the Ebola virus, which most likely jumped from fruit bats to humans, and Middle East Respiratory Syndrome (MERS) coronavirus, which jumped from camels to humans.

These differing outcomes led us to ask why only a subset of emergent viruses are able to establish transmissible infections in humans. In other words: what determines the likelihood of a virus’ human-to-human transmissibility?

We thought that biological factors such as the virus’ size, structure and mode of transmission might determine this likelihood. If we could show this to be the case, then identifying the determinative biological factors in emerging viruses could assist public health planning and resource-allocation decisions. New emerging viruses could be categorised according to whether they are likely to become transmissible among humans or to generate “dead-end spillover infections” like bird flu.

Previous work appeared to be either overly general or overly specific when it came to addressing the likelihood of sustained human-to-human transmissibility of emerging diseases. In the former camp, useful links had been highlighted between the likelihood of new diseases emerging and a variety of socio­economic factors, including sanitary conditions and political instability. Further, it had been suggested that collating data on the occurrence and distribution of emerging diseases could be used to identify geographical hotspots where new diseases would be likely to emerge in humans.

The limitation with these approaches was that although they might show where and when new diseases might emerge, they did not discriminate between likely dead-end or human-to-human-transmissible diseases.

Other work tended to be highly pathogen-specific, considering the number and origin of the mutations that would be necessary to allow adaptation to human hosts and then human-to-human transmissibility. From our perspective this type of work made it difficult to draw general conclusions about the likelihood of human-to-human transmission by other viruses.

What 200 Human Viruses Tell Us

Instead our research, published in the Proceedings of the National Academy of Sciences (, looked at the biological features displayed by viruses that had achieved sustained human transmission. We compiled and analysed a database of more than 200 human viruses, and assessed whether viruses exhibiting particular biological features were more often associated with sustained transmission among humans. The biological features we considered reflected key aspects of viral life history and ecology, such as:

  • host mortality rate;
  • whether the genome is comprised of DNA or RNA;
  • genome size;
  • whether the viral genome is broken up into distinct segments or exists as one complete segment;
  • the frequency of genomic recombination;
  • the presence or absence of an outer envelope;
  • the duration of infection; and
  • the mode of virus transmission.

We then set out to determine which of these biological features, singly or in combination, was most often associated with human-to-human transmission, and hence which biological attributes of viruses increased the likelihood of human-to-human transmissibility.

Our analysis revealed a number of strong associations between the likelihood of a virus’ human-to-human transmissibility and its biological features. In particular, we determined that viruses were more likely to be transmissible among humans if they:

  • caused low host mortality;
  • established long-term chronic infections;
  • are not segmented or enveloped; and, most importantly
  • not transmitted by vectors such as mosquitoes.

In contrast, the frequency of recombination and the size of the virus’ genome were less important predictors of transmission success. Strikingly, genome type (DNA or RNA) had essentially no predictive power.

Overall we identified many biological features that appear to determine the likelihood of inter-human viral transmissibility. This in turn enables general predictions to be made about whether an identified emergent virus will achieve human-to-human transmissibility.

However, our study does not enable us to predict how sustained this transmissibility might be, and this can be the difference between a full-blown pandemic or an isolated outbreak.

Our Results in Detail

Our results strongly indicate that the likelihood of human transmissibility decreases as the host mortality rate (i.e. virulence) increases. Although the relationship between virulence and transmission is complex, the notion that low host mortality will generally allow more time for inter-host transmission seems well-founded.

However, it must be noted that estimates of host mortality rate rely heavily on precise diagnosis and accurate reporting. The mortality rate may be vastly overstated in the case of rare viruses that are under-reported or viruses that can establish asymptomatic infections.

Our analysis also reveals that the length of time it takes for a virus to replicate within an individual human host – quantified as either “acute” (<4 weeks) or “chronic” (>4 weeks) – is an important parameter in determining whether a virus can evolve human-to-human transmission. Chronic viruses are more likely to be transmissible between humans because extended durations of infection increase the chances of secondary transmission to a new host.

Of the 69 vector-transmitted viruses in our list, only six were transmissible between humans. That vector-borne viruses are less likely to jump to a new host and successfully establish an infection is to be expected given the complexity of animal-to-human disease transmission cycles that involve invertebrate vectors and vertebrate hosts.

Remarkably, while sexual transmission of Zika virus has been reported between humans, this is the only vector-borne virus in our data set where onward human transmission may not involve the usual animal-to-human disease cycle. Humans are usually dead-end hosts for vector-borne viruses, presumably because viral loads are insufficient to allow onward transmission through a biting vector.

One of our more puzzling observations is that non-segmented viruses with non-segmented genomes seem to be more capable of transmission among humans than those with segmented genomes. This finding is novel; there is no obvious explanation for it.Indeed it’s possible that this is a confounding observation obscuring the true underlying drivers.

In this regard, we note that all the segmented viruses in our data set develop acute infections, which is itself associated with a decreased probability of human-to-human transmission. In addition, many DNA viruses establish a chronic infection and are never segmented.

On the other hand, it could be that non-segmented viruses have simpler genomes and therefore a simpler, and presumably quicker, replication process that may benefit host adaptation. However, this is far from certain and this question warrants further research.

Another notable finding is that viruses with a viral envelope were less likely to establish human-to-human transmission than non-enveloped viruses. It’s possible that non-enveloped viruses are more environmentally stable than their enveloped counter­parts as the glycoproteins and lipids that form the envelope are easily degradable and may decrease the probability of inter-host transmission through contact with exposed surfaces. Indeed, non-enveloped viruses are more resistant to common ethanol disinfectant.

Finally, prior research has speculated that the frequency at which viruses undergo genetic recombination may determine inter-human transmissibility of emerging viruses. However, we found that recombination rates have little predictive power in this regard. Although recombination has the potential to facilitate transmissibility by accelerating the rate at which advantageous genetic combinations are produced, frequent recombination will also break up beneficial genetic configurations.


Until recently, much research on new emerging diseases has attempted to reveal the processes leading to disease emergence in human populations rather than focusing on the subsequent transmissibility among humans. Our study, however, has revealed factors that can be used to predict the likelihood of transmissibility of emerging pathogens between humans. These factors may explain why some viruses are more readily transmitted among the human population than others.

By identifying the major biological features of successfully emerging viruses, our analysis can be used to generate broad predictions of the likelihood that a specific virus will achieve human-to-human transmission and epidemic spread.

Jemma Geoghegan is a Postdoctoral Research Fellow in the Charles Perkins Centre, Marie Bashir Institute for Infectious Disease & Biosecurity at The University of Sydney.