Genomelink is a new standard to know more about your DNA data. We help you understand your DNA identity with intuitive visualization and scientifically educational, entertaining content.API Genetics
The field of human genomics has exploded the past decade. Thanks to better, faster and cheaper technology, the price to sequence a human genome (basically to extract the “DNA code” of a person) dropped from $100M in 2001 to 1300$ in 2015😲, opening the door to a myriad of new applications.
Since sequencing a genome became so cheap, researchers increasingly conduct studies comparing the genome - as humans we share 99.5% of the same DNA code, the 0.5% left is what makes each of us unique - of specific groups of people, to see what are the DNA characteristics they shared. For example, they sequence the genome of people suffering from Alzheimer to detect the specific DNA configuration they have in common, inferring that people with a similar configuration are possibility more at risk to suffer from Alzheimer.
From there, you can separate healthcare genomics startups, which are operating in regulated environments for “serious” healthcare applications, and consumer genomics startups, which are leveraging genomics for genealogy (like 23me or Ancestry), wellness and even entertainment (like this pack which recommends wine based on your DNA :-)). Needless to say that the latter doesn’t operate under the same regulation, nor the same scientific requirements as the healthcare startups.
In the memo of the week we’ll explore Genomelink which aims to provide the infrastructure for consumer genomics companies.
I want to focus on the B2B product here. As I mentioned quickly in the intro, there’s an explosion of consumer genomics startups (for genealogy, wellness and entertainment). Many of these startups share the same concept: you order a DNA kit from their website that you mail back (generally with your saliva). They then send your kit to a lab which will sequence part of your DNA. Once they have your genome (or part of it at least) they compare it with what the available studies say to provide you with “personalized recommendations” based on your genetic (whether it’s for a diet, personalized supplements or even a sport training program).
What is fascinating to me, is that actually very few of these startups have the resources to hire their own researchers and conduct their own research. Plenty are pure intermediaries: paying labs for sequencing the DNA of their customers and using studies that they haven’t conducted to provide the recommendations :-)
And Genomelink approach is to make the “recommendation component” available “as a service”: they aggregate data coming from many studies and let developers retrieve custom genetic insights based on their users' genomic data with a simple API call. Fascinating (both for good and bad reasons).
Remember when I explained that an increasing number of genomics studies are conducted by researchers to link specific DNA configurations to particular diseases, physical or personality traits? Genomelink basically aggregates this data to make it available through two different products:
I cannot finish this memo without mentioning the importance of data here. Obviously being able to generate proprietary data is crucial to offer unique insights and build defensibility. This is why the biggest consumer genomics startups, such as 23andme or Ancestry, have raised hundreds of millions of dollars. They do have teams of researchers working in house and own their labs. Being an aggregators or using public genomic research/data will only take you so far in this industry.
Let’s be honest, the majority of consumer genomics startups are currently borderline BS (not speaking about Genomelinks in particular, they are more on the educational side, but about the wellness and entertainment genomics ones). In the US they are barely regulated. Even if you pay for the most well funded players, you can get different results based on the same genome (this article explains why), showing that we are not there yet in terms of reliability. But the field is progressing fast, and we’ll probably get there in some years.