pademic predictions, nature perspective

 Fluleap 

Could an algorithm predict the next pandemic? (nature.com)

, and the virus’s genetic sequence was quickly uploaded to the genetic data repository GISAID. For Colin Carlson, a biologist at Georgetown University in Washington DC, it presented an opportunity. “I immediately thought, ‘I want to run this through FluLeap’,” he says.

Researchers estimate that around 1% of the mammalian viruses on the planet have been identified1, 

1. Carlson, C. J. et al. Phil. Trans. R. Soc. Lond. B 376, 20200358 (2021).

the names and affiliations mentioned are as follows:

1. Colin Carlson - Biologist at Georgetown University in Washington DC, also the director of the Viral Emergence Research Initiative (Verena).

2. Kevin Olival - Ecologist and study leader at the EcoHealth Alliance in New York City.

3. Jonna Mazet - Epidemiologist at the University of California, Davis, and director of the PREDICT project.

4. Edward Holmes - Virologist at the University of Sydney in Australia.

5. Jemma Geoghegan - Virologist at the University of Otago in New Zealand.

6. Sara Sawyer - Virologist at the University of Colorado, Boulder.

7. Nardus Mollentze - Computational Virologist at the University of Glasgow, UK, and collaborator with Verena researchers.

The article also mentions the following organizations and projects:

1. GISAID - Genetic data repository

2. PREDICT project - A US$200-million project funded by the US Agency for International Development (USAID)

3. Global Virome Project (GVP) - Proposed project in 2016, currently a non-profit organization

4. Discovery and Exploration of Emerging Pathogens — Viral Zoonoses (DEEP VZ) - Project launched by USAID in October 2021

5. Viral Emergence Research Initiative (Verena) - A consortium of researchers seeking to develop and improve zoonotic prediction models.

Some key important points in the article are:

1. In February 2021, seven Russian poultry-farm workers were infected with the H5N8 avian influenza, a subtype of bird flu that had not previously been known to infect humans.

2. Colin Carlson, a biologist at Georgetown University, used the machine-learning algorithm FluLeap to classify the H5N8 virus as human with 99.7% confidence, suggesting the model may have inferred a biological signature of compatibility with humans.

3. The zoonotic process of viruses jumping from wildlife to people causes most pandemics. Climate change and human encroachment on animal habitats increase the frequency of these events.

4. Machine learning could help identify the viruses most likely to spill over from animals to people and cause future pandemics.

5. Researchers have used statistical models and machine learning to predict aspects of disease emergence, such as global hotspots, likely animal hosts, or the ability of a particular virus to infect humans.

6. PREDICT, a US$200-million project funded by the US Agency for International Development (USAID), identified 949 new viruses in samples from wildlife, livestock, and people in 34 countries.

7. Machine learning models could be used to flag high-priority targets for further investigation, helping to triage newly discovered viruses and guiding the development of vaccines and therapeutics.

8. Improving the data used by artificial intelligence algorithms is essential to their success, and it requires global cooperation, open data sharing, and adherence to data standards.

9. Overcoming political, cultural, and ethical obstacles is crucial for effective data sharing and collaboration, which can help build trust and benefit countries that share genetic data.