As many of you will remember, our research began in the summer of 2020. Spain, the driving force behind European tourism, was preparing to welcome thousands of citizens after months of lockdown. At that time, concern was at its peak: how could we quickly and massively control the entry of the virus at borders, airports and tourist destinations?

We all know that dogs are excellent detectors, which is why they were tried from the outset. However, in the early stages, training was based on using masks worn by patients, which presented two major risks:

• Biosafety: We did not know for sure whether exhaled breath or saliva could infect the handlers or the dogs.
• False positives/negatives: When using a mask, the dog could memorise the individual smell of that infected person (their diet, age, gender, race or perfume) instead of the common smell of the virus, reducing the universality of the detection.

Motivated by canine professionals who needed real tools to train safely and effectively, we set ourselves a challenge: Can we ‘capture’ the smell of COVID-19 in humans to train dogs in a 100% safe and effective way?

📊 The Scientific Rigour Behind the Sense of Smell

Yesterday, we published our study entitled  “Untargeted Metabolomics and Multivariate Data Processing to Reveal SARS-CoV-2 Specific VOCs for Canine Biodetection” in the journal Chemosensors.

Thanks to the involvement of our researchers, especially Eider Larrañaga and Diego Pardina, we have managed to decode the volatile signature of the virus in sweat samples.

We applied untargeted metabolomics and machine learning models to find patterns. This work provides a scientific basis for what dogs naturally detect; we finally know what dogs associate with detecting COVID-19.

• Identification of Biomarkers: Among the hundreds of compounds emitted by the human body, we identified six key Volatile Organic Compounds (VOCs) that are consistently present in positive cases.
• Universality: These markers appear regardless of age, sex, or virus variant.
• Accuracy: The model achieved 89% specificity, validating that the ‘characteristic aroma’ we have isolated is the correct one.

🐕 What does this mean for the K9 world?

It means we no longer need dangerous or individual biological samples. By identifying these specific biomarkers, at Auziker we can create universal training aids.

The dog learns the ‘pure scent’ of COVID-19 safely, becoming a much more selective and reliable biosensor for mass screening. The smell does not depend on how much solid there is, but on how many molecules pass into the air and how they interact with the dog’s olfactory receptors.

Hopefully, these baits will no longer be necessary for this pandemic, but this work has opened up a revolutionary working methodology, which has allowed us to prepare aids such as Hilotz for the detection of corpses in the early stages of decomposition. This paper is the result of our commitment: to provide the scientific basis for manufacturing K9 training aids for future biological threats.