DNA & AI: the perfect match for personalisation


DeoxyriboNucleic Acid, or DNA, is basically a molecular blueprint for any living thing. Our genes (we have about 20.000 to 25.000 of them) are portions of DNA that can code for a particular characteristic or trait, such as the shape of our ears or the colour of our eyes.
The unique nature of DNA is typically what enables forensic scientists to help catch criminals. That’s not particularly new, of course. But one of the most recent developments that I’m very excited about is how we could couple DNA with AI algorithms for purposes of extreme personalisation in the consumer world.

Commercial genetic testingThis chart right illustrates how commercial genetic testing is growing in popularity, largely based on the success of personal genomics and biotech companies like AncestryDNA and 23andMe. Estimates put together by the MIT Technology Review, show that the number of 26 million testing kits sold could go through the roof and reach 100 million within the next two years. The possibility of using DNA samples to identify health risks, to track your ancestry, but particularly to detect your personal preferences, presents huge opportunities to improve customer experiences.


To show you all the amazing possibilities of such an approach, I gathered some use cases as an inspiration for your own company.

WellbeingDNA24's daily companion app
Using LifeNome’s AI & Genomics algorithm, DNA24 provides a ‘daily companion app’ that – based on your DNA profile and personality traits – tells you which exercises to do throughout the day and plans personalised meals for you.

BeautyAI also uses Lifenome’s algorithm, to understand your unique skincare needs and recommend the best products.
Similarly, L’Oréal is partnering with UBiome, a microbial genomics company that claims to have the largest microbiome database in the world with over 250,000 samples. Consumers can submit their cheek-swab sample to have their skin health determined ‘based on the ecosystem of microbes of bacteria, viruses, fungi and other microorganisms’. L’Oréal uses these results to offer more personalized product choices.

Currently, most available drugs are “one size fits all” and it is hard to predict who will benefit from a medication, who will not respond to it at all, and who will experience negative side effects. A relatively new science called “Pharmacogenomics” combines Pharmacology, the
science of how drugs work, with Genomics, the science of the human genome.Precision medicine - tailored doses based on genes
It studies the way genes affect a person’s response to drugs, and big pharma companies are using this knowledge to develop safe medications and doses tailored to a person’s genetic makeup. This ‘precision medicine’ is still in its infancy, but soon doctors will be able to use information about your genetic makeup to choose – using AI-powered algorithms – the “best-fit” drug and drug dose, with the greatest chance of helping you.

Personalised nutrition may be taken to the next level by a company called Zoe. Still in its prototype phase, they are aiming at creating an app that would offer users individualized nutrition advice about how to eat and, ultimately, how their bodies might respond to foods they have not yet tried. The goal of having ‘diets for one’ is drawing nearer.

DNA-driven dinners with DNAFit and VitaMojoAt the London-based take-away restaurant Vita Mojo, the mission of personalised nutrition is already being realised. They partnered with fitness genomics company DNAFit to offer “DNA-driven dinners”. Based on analysis of cheek swab examples, you will find out whether you are sensitive to carbohydrates, salt or saturated fat, and whether you have lactose or gluten intolerance. Next to that, your individual needs for antioxidants and vitamin are also calculated. This DNAFit service is then combined with the Vita Mojo algorithm to recommend meals that consider your individual DNA, health and fitness goals and taste preferences.

But what to think of DNA matchmaking? This may prove a welcome solution for the lonely hearts out there. DNA matchmaking services are beginning to crop up in Japan (Gene Partner) and Singapore (Genemate), and some governments even see it as a fix for low birth rates. HLA genes are being analysed and combined with for example hobbies and interests to ‘find soulmates’ and even ‘create long-lasting marriages’.
GenemateUS-based DNA Romance is a free online dating site for people looking to find genuine relationships based on chemistry and personality compatibility. They use AI to decipher the essential elements behind the “scent of love” and the ideal personality combinations for successful relationships full of love. Upload your DNA, fill in a personality test and within 5 minutes you can start swiping your matches. Now, as fun as this might sound, there might very well be some ethical consequences to such an approach. What if these types of companies would start selecting on race? Only time will tell how ethical and successful these approaches will prove to be.

Music and beer
Some applications of DNA and AI appear to be rather gimmicky and less scientific.

Spotify is known for its personal recommendations. In fact, the suggestions for other playlists or the introduction of artists previously unknown to you based on your ‘listening behaviour’, has largely contributed to the company’s success. They claim that, using your AncestryDNA results, they can create a unique, ‘eclectic’ playlist inspired by your origins. Hm, this sounds a bit like ‘music based on race’ doesn’t it?

And then there’s personalisation of beer glasses based on users’ DNA. Japanese brewing company Suntory creates 3D printed glasses, with the capacity determined by the alcohol tolerance, the top diameter by the sensitivity to malt aroma and the rim thickness by sensitivity to hops bitterness. A nice gimmick, but no proof of it making real sense and adding to the customer experience.Suntory - personalised beer glasses



What other types of personalisation based on DNA can we expect? Maybe ‘DNA Residence’ to suggest the best place for you to live based on your health condition, allergies and sensitivity to respiratory tract infections. Or ‘DNA Pet’, an algorithm to find the right type of pet for you. Specific fur allergies would rule out certain types of cats, life expectancy could withhold you to get a parrot (they do get very old), and future health condition may suggest choosing a fish – not in need of exercise – over a dog.

The potential applications of DNA-based AI are endless and with the exponential growth of data collected, new recommendation engines will quickly be built.

Of course, privacy and ethical issues may very well be lurking around the corner. The direct-to-consumer DNA kit 23andMe allows individuals to assess their potential health risks by predicting genetic illnesses. What if this information falls into the hands of health insurance or life insurance companies? Should they in fact be allowed to request these test results like these?

As consumers, we love personalisation and we are often willing to pay more for an extremely tailored product or service, even with our personal data. And using the same username and password in different settings is of no big concern to us, as we think we are safe because we can always change those. But our own DNA, unlike usernames or passwords, can never be changed. Once it is in the wrong hands, it cannot be ‘undone’.