Predict your next investment

Corporation
HEALTHCARE | Disease Diagnosis
23andme.com

See what CB Insights has to offer

Founded Year

2006

Stage

PIPE | IPO

Total Raised

$877.69M

Revenue

$0000 

About 23andMe

23andMe is a personal genetics company dedicated to helping individuals understand genetic information through DNA analysis technologies and web-based interactive tools. The company's Personal Genome Service enables individuals to gain deeper insights into ancestry and inherited traits. The vision for 23andMe is to personalize healthcare by making and supporting meaningful discoveries through genetic research.

23andMe Headquarter Location

233 N. Mathilda Avenue

Sunnyvale, California, 94086,

United States

650-938-6300

Latest 23andMe News

Mohammed is Palestinian. Why does 23andMe think he’s Egyptian?

Sep 15, 2021

Genetic ancestry tests such as 23andMe attempt to place individuals into neatly-defined country groups. But people from Palestine are finding their heritage erased When Mohammed Ahmad took a 23andMe genetic ancestry test in 2019 he wanted to understand what his DNA said about his family history. Ahmad, aged 23, had always been fascinated with the past and particularly the history of the Middle East, where his parents were born. What Ahmad wasn’t looking for, however, was a cultural identity. He already knew who he was and how he’d always identify: Palestinian. Ahmad was born and raised in the US but knows his family history well. His parents migrated after they met in Asira al-Qibliya; a village of around 2,300 people outside the city of Nablus in the West Bank. This village was home to not only his father, but his grandfather and his grandfather’s father. Because of his family’s deep roots in the West Bank, Ahmad was not surprised when he opened the letter from 23andMe containing his results: 100 per cent North African and Arabian. But not too long after taking the test, the story of Ahmad’s genetic ancestry started to rewrite itself. Email notifications in October 2019 and June 2021 told Ahmad that updates to 23andMe’s system of genetic categorisation had changed his results. In the most recent update, Ahmad’s DNA is broken down into the following percentages: 49.4 per cent Lebanese, 38.1 per cent Egyptian, 2.7 per cent Peninsular Arabian, 3.7 per cent Iranian, Caucasian and Mesopotamian, plus small percentages from other regions. What stuck out to Ahmad was the glaring absence of Palestine in all of these updates. “As a Palestinian, your identity is always questioned,” he says, “I feel like finding out about our DNA and finding out that, ‘yes, we are from this land’ is a big thing.” While some customers may have been satisfied with the 23andMe’s more granular results, some Palestinians like Ahmad are unhappy with being labelled under other countries in the Levant region (which includes Syria, Jordan, Lebanon, Israel and Palestine) as well as states in the Arabian Peninsula (which include Kuwait, Oman, Qatar, Saudi Arabia, Bahrain, Yemen and the United Arab Emirates). “I found it interesting how they name every country surrounding Palestine but Palestine,” Ahmad says, “It was disheartening to find out that Palestine wasn’t being mentioned.” Neither Palestine or Israel is listed as a possible country of ancestry in the company’s test results. “At the end of the day cities like Jerusalem, Jaffa,  Acre, Jericho and Hebron have had inhabitants for thousands of years,” says Amber-Rose Kedem, another 23andMe customer. When Kedem first took the test in 2019, her two largest percentages were Peninsular Arab (25 per cent) and Levantine (15 per cent). But when she decided to check her updated reports earlier in this year, Kedem was categorised as 37 per cent Peninsular Arab and only 0.8 per cent Levantine. “I see it as purposeful erasure,” she says. In 2020, an online petition was started on Change.org titled “Recognise Palestinians – 23andMe”. The petition author argues that 23andMe has cast Palestinians as “genetically stateless” by excluding the region from its reports. It ends: “We are Palestinians. We exist. We are still here and we will never go away or be erased from history.” Over 2,400 people have signed the online petition. Updates, such as the ones that have affected Ahmad and Kedem’s results, came after 23andMe entered more results into its database, allowing the company to narrow down its customers’ ancestral ties. But the company’s updates have failed to represent people like Ahmad who have Palestinian ancestry. Samantha Esselmann, a 23andMe product scientist, says that there are two main reasons for this: firstly, Palestinians did not form their own distinct genetic cluster, and secondly, even if they did, the changing borders of Palestine and Israel made accurate labelling of that cluster too difficult. Categorising people’s genetic ancestry is notoriously difficult and subject to all kinds of biases, but the technique that 23andMe uses hinges around reference populations: a group of people whose genetic similarities represent a country or region that a new customer can then be compared to. The company uses samples from public databases as well as samples from their customers to form these reference populations. Using something called “principle component analysis” (PCA), 23andMe studies the genetic variation across these samples. What scientists tend to see is that people with geographically close ancestors tend to be more genetically similar. People who were born in Sweden may be more genetically similar to one another than people who were born in Italy, for instance. And if we were to plot people as data points, those who come from the same places tend to form “clusters”. If 23andMe notes that a cluster has many individuals who all state that their grandparents came from, say, Country X, then the company may label that cluster “Country X” and the people within that cluster become X’s reference population. This methodology inherently relies on people identifying where their recent ancestors came from to help 23andMe label clusters appropriately. This becomes difficult if recent ancestors came from states with fluctuating borders. “If ‘Palestine’ and ‘Israel’ mean different things to different people, depending on when your grandparents were born and where, exactly, they were born, then it becomes difficult to define clear Palestine and Israel reference populations using our existing ‘Recent Ancestor Location’ methodology,” says Esselmann. New customers are compared against these reference populations, so the percentages that customers receive reflect their genetic similarities to 23andMe’s other customers. This may be different from how many customers interpret their results. A 25 per cent Italian result is often taken to mean that a grandparent is Italian. However, a more accurate interpretation would be that the customer shares a 25 per cent similarity with 23andMe’s Italian reference population, suggesting some form of Italian ancestry. Esselmann compares the company’s process to paint matching. A person’s DNA can be imagined as a paint sample, she says. The DNA is broken up into thousands of segments which are then compared to a number of reference populations. In 2008, these reference populations were relatively general: European, Asian, African and so on. “If you have four easily differentiable colours like green, yellow, red and blue, you can say, well, this segment [of a person’s sample] is not this exact colour of green, but it best matches this green bucket,” Esselmann says, “That would be equivalent to saying I have Italian ancestry, and it best matches the European bucket.” But over time, the company’s reference populations have grown. Customers are asked to identify the countries where their grandparents have spent most of their lives, and their samples can be entered into 23andMe’s database. These more granular reference populations are like shades of green or blue that can now be matched to a new customer’s DNA. The company’s analysis will match segments of DNA to the growing number of different reference populations (hence customers receive percentages with their results). That’s why updates lead to different test results. The company has updated its reference populations between two to three times a year since 2018. Marguerite Dabaie, a Palestinian-American artist, was first identified as 85 per cent European (50 per cent British and Irish, and 35 per cent Italian) and only 15 per cent North African or Arabian when she first took the test in 2016. But her most recent report categorises her as 50 per cent European – all of which is British and Irish – and 50 per cent Western Asian and North African. “I get that it’s a work in progress but on the other hand, I don’t even see them marking Palestinian DNA as a possibility,” Dabaie says. Even if Palestine was a fully recognised state, Esselmann says Palestinians do not form a distinct cluster under the company’s current analysis. “The individuals who said they were from Palestine, in our analysis, were clustering with people who were from Jordan, Lebanon, Syria etc,” Esselmann says, “Even though we had Palestinian samples, there was no distinct cluster that we were able to distinguish with confidence that included only people from Palestine.” However, such processes, because they attempt to find such minute genetic variations between people, are highly sensitive to sample quality as well as the granularity of the analysis. Aylwyn Scally, a genetics researcher at the University of Cambridge, says that clusters are dependent on the scale at which you look at genetic variation. For instance, data points from Asia may be clustered together, but zoom in and people from different states may form clusters, and at an even smaller level, people from different tribes may form clusters. If you look hard enough, there may well be a Palestinian cluster. Another thing that impacts clustering is sample quantity. “Having more actual representatives of people in this region in their database gives them greater power to distinguish between groups,” Scally says. “The more data you have, the more you're able to discern patterns in one that are present in some people and not others.” Scally says there’s nothing stopping 23andMe saying that its Levantine cluster includes people from Lebanon and Palestine for example. Clustering analysis is a blend of creation and discovery. The human genome is complex and doesn’t form neat categories. Genetic ancestry companies have to label their results in a way that glosses over these complexities in order to create simple but meaningful results for customers. After being dissatisfied with his results, Ahmad took a test with AncestryDNA which categorised him as 98 per cent Palestinian. “The AncestryDNA results impacted me greatly – it actually made me really happy to finally see the word ‘Palestine’,” Mohammed says, “When we’re looking at DNA, we’re looking at history and when we look at history, we need to mention ethnic identities that are on the brink of being wiped out. We need to mention Palestine.” More great stories from WIRED

Predict your next investment

The CB Insights tech market intelligence platform analyzes millions of data points on venture capital, startups, patents , partnerships and news mentions to help you see tomorrow's opportunities, today.

Expert Collections containing 23andMe

Expert Collections are analyst-curated lists that highlight the companies you need to know in the most important technology spaces.

23andMe is included in 6 Expert Collections, including Genomics.

G

Genomics

823 items

Companies involved in the capture, sequencing, and/or analysis of genetic data

S

Smart Money VCs (2017-2019)

6,297 items

We crunched the data to identify the 24 VC firms with the best combination of portfolio valuations and investment outcomes.

G

Google In Healthcare

61 items

The healthcare companies and initiatives that Google has invested in or is a part of

H

HLTH

199 items

HLTH is a healthcare event bringing together startups and large companies from pharma, health insurance, business intelligence, and more to discuss the shifting landscape of healthcare

T

Tech IPO Pipeline 2019

286 items

D

Digital Health 150 (2019)

150 items

2019's cohort of the most promising digital health startups transforming the healthcare industry

23andMe Patents

23andMe has filed 82 patents.

The 3 most popular patent topics include:

  • Genetics
  • Molecular biology
  • DNA
patents chart

Application Date

Grant Date

Title

Related Topics

Status

2/2/2021

9/14/2021

Graphical user interface elements, Graphical user interfaces, Graphical control elements, Graphical user interface testing, User interfaces

Grant

00/00/0000

00/00/0000

Subscribe to see more

Subscribe to see more

Subscribe to see more

00/00/0000

00/00/0000

Subscribe to see more

Subscribe to see more

Subscribe to see more

00/00/0000

00/00/0000

Subscribe to see more

Subscribe to see more

Subscribe to see more

00/00/0000

00/00/0000

Subscribe to see more

Subscribe to see more

Subscribe to see more

Application Date

2/2/2021

00/00/0000

00/00/0000

00/00/0000

00/00/0000

Grant Date

9/14/2021

00/00/0000

00/00/0000

00/00/0000

00/00/0000

Title

Subscribe to see more

Subscribe to see more

Subscribe to see more

Subscribe to see more

Related Topics

Graphical user interface elements, Graphical user interfaces, Graphical control elements, Graphical user interface testing, User interfaces

Subscribe to see more

Subscribe to see more

Subscribe to see more

Subscribe to see more

Status

Grant

Subscribe to see more

Subscribe to see more

Subscribe to see more

Subscribe to see more

23andMe Web Traffic

Rank
Page Views per User (PVPU)
Page Views per Million (PVPM)
Reach per Million (RPM)
CBI Logo

23andMe Rank

CB Insights uses Cookies

CBI websites generally use certain cookies to enable better interactions with our sites and services. Use of these cookies, which may be stored on your device, permits us to improve and customize your experience. You can read more about your cookie choices at our privacy policy here. By continuing to use this site you are consenting to these choices.