A test to predict wine preferences

Wednesday, 4. October 2017 - 17:00

A preference for blackberries or a dislike of coffee: what does it tell you about the wines you enjoy? How much is nurture versus nature is still uncertain, but the further scientists dig into DNA, the more they understand how much taste preferences could be defined by genes.

When I first met the Healdsburg, California-based Vinome team, I was doubtful if DNA could help consumers better understand their wine preferences. Vinome’s system of taste profiling involves taking a short flavour quiz, and then mailing in a test tube of saliva, from which DNA will be sequenced.

Vinome, launched this August, works in partnership with Silicon Valley-based internet platform Helix, a genome sequencing company that has developed a genetics app store. Customers pay $80.00 for gene sequencing, which can then be used in tandem with their other products, to offer insights into things like ideal sleep patterns, ancestry or health and nutrition information. Many of those applications cost approximately $30.00 each.

Brave New World

Every wine sales site is looking for a magic bullet that will help them determine what wines their customers would like to buy. However, the Vinome approach is unique in using genetic data. The company then offers new customers wine made by its partner wineries, which it claims are boutique bottles perfectly matched to the customers’ genetic preferences. So it is also functioning as a direct-to-consumer sales vehicle for domestic wineries, most of which are based in California.

The current eight taste profiles that Vinome currently offers are a little too broad; however, the company hopes to both expand them and hone them down once it gets more genetic data.

My test results were mostly spot on, with my major profile being “The Big Bold,” meaning “coffee, chocolate, smoke, tannins and leather.” My secondary flavor profile was the “Jam Dunk,” which is focused on big, jammy flavors. I like some of these flavors, but in moderation and with balance. The results are indicated pictorially, with different ingredients pictured inside bubbles. The larger the circle, the more you like that food taste.

I also came out as a big fan of mushrooms, blackberry and smoke in wines, which I am.  Co-founder and lead scientist Sarah Riordan, one of the application’s co-founders, also cautioned that those are flavour profiles that consumers may appreciate the smell of in wine more than the actual food itself. So now it makes sense to me that while most chocolate doesn’t appeal to me, a note of chocolate aroma in some wines is appealing.

The background

To develop the service, the Vinome team surveyed 541 people, asking them for information about their food and wine preferences and demographic information. Participants were then offered 12 wines and asked whether they liked them, and also if they could detect previously defined flavours. Each participant had their DNA sequenced, and Vinome looked at how genetic variants, known to affect particular taste preferences, matched their answers. Since then, another thousand people have been sequenced, according to founder Ronnie Andrews.

Everyone who signs up for Vinome receives a representation of ten taste genes, showing which allele – an alternative version of a gene – they have. For example, people who have particular alleles for the TASR50 gene find that cilantro tastes unpleasantly soapy; I don’t have that allele and it’s true that I can take or leave the herb. 

It gets even more interesting when it comes to other genes such as TASR20. Those who have at least one G allele can experience quinine, the bitter flavour in tonic, as extremely bitter. That sour note is an element I appreciate in tonic. Andrews added that there was no right or wrong answer as to how one perceived specific flavours, but the DNA sequencing offers an insight into how these alleles might affect one’s appreciation of certain wines. Again, my results suggest that I can better taste a bitter chemical found in Brussel sprouts and other cruciferous vegetables, which are among my favorite foods in great part for just that reason. So my DNA-based flavor profile was mostly accurate. One of the few off calls was the suggestion that I have a sweet tooth. I don’t ordinarily eat processed sugar, but don’t mind the occasional caramel note in wine.

Once potential customers are sequenced the Vinome website then directs them to three-, six- or 12-pack wine club offerings delivered quarterly which cost from $149.00 to $599.00. Wines can also be purchased individually according to taste profile under the “Shop my Vinome,” drop-down menu. The bulk of those suggested for the “Big Bold,” flavour profile were California Cabernet Sauvignons, which I generally find to be too high alcohol. Several of the suggestions for my secondary taste profile—the “Jam Dunk”—were a better stylistic fit, such as the Consilience 2014 Pinot Noir from Santa Barbara. I suspect that if Vinome branches out to selling wines from other, cooler-climate areas of the world the application might be better able to offer me wines more suitable to my palate.

However, the genetic basis of taste is not well understood and other scientists have been sceptical of Vinome’s claims. Some of the services offered on the Internet may be based on genetics, but still may not provide valid information, cautions Jonathan Berg, an associate professor in the department of genetics at the University of North Carolina at Chapel Hill. He adds that it is often difficult to tell useful from misleading information and that it will be difficult “for consumers to know what is valid and not valid”.  Critics aside, I found the exercise fascinating and mostly accurate.

Liza B. Zimmerman