A brief call towards investigating the potential of volatolomics for high-throughput and non-invasive disease detection.

[Thanks to Nikhil Lal, Karolina Sulich, and Oliver Crook for reviewing and providing helpful comments]

Massive multiplexed nucleic acid detection and metagenomic sequencing are crucial tools for an early warning system as well as for continued monitoring of Global Catastrophic Biological Risks (GCBRs). However, such tools are not best suited for the kind of high-throughput and minimal or non-invasive screening that is crucial in both early and later stages of a pandemic to prevent transmission and exponential growth. The 2021 Apollo Program Report by the United States’ Bipartisan Commission on Biodefense accordingly notes the need to invest in technologies such as  - “New sensing capabilities … such as non-invasive volatolomics (the detection of volatile compounds emitted by an individual) … could permit constant passive monitoring of markers of infection without interfering with or inconveniencing our daily lives. Furthermore, non-invasive and minimally-invasive detection techniques could provide avenues to monitor high-risk, high-concern, and sentinel populations for infections, without disrupting daily life.” (emphasis mine).

The utility of volatolomics and its complementarity with other technologies is best illustrated by the use-case in public transit systems, although the same logic applies to other public spaces such as grocery stores, workspaces, stadiums, schools, nursing homes and prisons. The ability to passively and continuously screen every single person as they walk through a check-point and ascertain with a high degree of accuracy whether they harbour a particular infectious agent would be a huge asset at international hubs (preventing the start of a pandemic) and domestic hubs (restricting outbreaks to local nodes). The convenience of olfaction’s ‘user interface’ in disease detection - continuous, passive, and real-time - is highly valuable on its own, and also allows for a tiered screening system by enabling the triaging of other diagnostics such as CARMEN-Cas9.

A compelling case can be made for the foundations of such volatolomic technologies to already exist. Gas chromatography mass spectrometry (GC-MS) has undergone rapid improvements, and handheld ‘e-noses’ have been developed by multiple teams at universities such as MIT and Weizmann, as well as a plethora of start-ups, with advances in AI enabling inferences that was hitherto unavailable. New technologies that utilise hybrid ‘bioelectronic’ systems also show great promise by utilising biological receptors interfaced with artificial neural networks - e.g. Koniku (which has reportedly collaborated with Airbus) and Scentian Bio (https://www.scentianbio.com/). Strikingly, the most robust biodetection sensor remains the humble domestic dog. Biodetection dogs were employed in all 6 permanently-inhabited continents during the Covid-19 pandemic, with studies indicating that dogs outperformed ‘gold-standard’ RT-PCR (Hag-Ali, 2021). John Hopkins and UPenn researchers sketched out the case for the promise of disease detection dogs during future pandemics (Otto, 2021), and Nature covered the field in a 2022 report. Advances in the interdisciplinary working dog field, such as genetics and brain-computer interfaces, will likely further augment our abilities to work with dogs, and canine-in-the-loop automation techniques are likely to provide refinement for machine and hybrid systems.

Despite their immense promise, a thorough review, map or synthesis of volatolomics in augmenting biosecurity does not yet appear to exist. Moreover, discussion on volatolomics appears absent in the effective altruism/x-risk community (only two mentions on the EA Forum both simply highlighting the Apollo report). While technological advancement is likely to continue without national/philanthropic funding, as incentives for startups are high given the lucrative medical screening space, effective deployment for state and international biosecurity will require extensive coordination, policy implementations, and cooperation amongst industry, academia, and state actors, something that may be unlikely to occur without explicit intervention. The nature of volatolomics technologies also raises critical ethical questions regarding data privacy and ownership, that would be crucial to resolve before any wide-spread deployment. It is especially worth noting that volatolomic technologies largely remains a FDA classification grey zone and there is incomplete understanding on how HIPAA compliance will be achieved with the unique features of olfactory data.

The promise of volatolomics for biosecurity coupled with the specific type of work required to actualize its potential would appear to make volatolomics x biosecurity a compelling case for effective altruism to focus on. A natural starting point readily emerges from the fact that Osmocosm, a premier machine olfaction conference based at MIT, will take place the week before EA Global Boston, this October 2023. 

Open questions

What are the main bottlenecks to deploying volatolomic technologies in real-world biosecurity settings? Are they based on technological, regulatory or pragmatic challenges?

How much utility can volatolomic technologies realistically deliver during different stages of a GBCR, given innate constraints to the technology, such as the likely need to train noses on a novel disease volatilome?

Who are the major stakeholders in this space? Are there feasible frameworks that would amplify collaboration and cooperation between them?

Resources

Two years of talks freely available on Osmocosm  -  https://www.osmocosm.org/ 

References

1. Bipartisan Commission on Biodefense. The Apollo Program for Biodefense: Winning the Race Against Biological Threats. Bipartisan Commission on Biodefense. Washington, DC: January 2021 - https://biodefensecommission.org/reports/the-apollo-program-for-biodefense-winning-the-race-against-biological-threats/  
2. Toward a disease-sniffing device that rivals a dog’s nose, MIT News - https://news.mit.edu/2021/disease-detection-device-dogs-0217 
3. A sniff test for coronavirus? - Weizmann Compass - https://www.weizmann.ac.il/WeizmannCompass/sections/briefs/a-sniff-test-for-coronavirus 
4. Airbus and Koniku Inc. embark on disruptive biotechnology solutions for aviation security operations - Airbus Press Release - https://www.airbus.com/en/newsroom/press-releases/2020-05-airbus-and-koniku-inc-embark-on-disruptive-biotechnology-solutions 
5. Scientian Bio - https://www.scentianbio.com/ 
6. Hag-Ali, M., AlShamsi, A.S., Boeijen, L. et al. The detection dogs test is more sensitive than real-time PCR in screening for SARS-CoV-2. Commun Biol 4, 686 (2021). https://doi.org/10.1038/s42003-021-02232-9 
7. Otto, Cynthia M., et al. "The promise of disease detection dogs in pandemic response: lessons learned from COVID-19." Disaster Medicine and Public Health Preparedness 17 (2023): e20. https://doi.org/10.1017/dmp.2021.183 
8. The dogs learning to sniff out disease, Nature, doi: https://doi.org/10.1038/d41586-022-01629-8