Hacker News Reader

Comment by wenc | Hacker News Reader

I see the word DNA being bandied around here. I wonder if folks here understand the difference between "genotyping" (what 23andme does) and full "sequencing"? [1]

Genotyping only returns variants of interest (SNPs) which are specified up front -- not the full sequence. It's a smaller dataset.

Sequencing gives you the full genome -- which is currently too expensive to do at consumer scale (though prices are falling). 23andme does not do this. This means the type of genetic research that 23andme can do is somewhat limited.

[1] https://customercare.23andme.com/hc/en-us/articles/202904600...

View on HN

Is genotyping data that useful for drug development?

From my intuition, genotyping data (where you specifically look for already known gene variants) is per definition barely useful for making "breakthrough discoveries". If that were really the case I would expect a much higher price point than $300mil.

It absolutely is extremely useful. Your genetic predispositions toward disease, treatment response, anything really, can be due to a combination of these known variants. Tracing your ancestry is done with these variants. These are regions that are commonly variable in the population.

https://en.wikipedia.org/wiki/Single-nucleotide_polymorphism

> which is currently too expensive to do at consumer scale

There are companies that do that for $200, sequencing is the new genotyping!

(which I hope will encourage more people to donate their data to open source databases, making the sale of this kind of data unprofitable)

The cheapest I could find in 2019 was $600 by Veritas Genetics for next-gen sequencing. 23andMe is SNP testing as OP said, which is cheaper.

http://arep.med.harvard.edu/gmc/genome_services.html Here are some more options! Some usually have holiday deals.

https://us.dantelabs.com/

oh shoot. they had a deal last year, $200 for 30x coverage

These days, thanks to large haplotypic reference panels, genotyping with an adequately dense array is more than sufficient to impute almost everything else with extremely high accuracy.* You might miss a few rare variants but not much else.

*Assuming your ancestry is from a well studied population e.g. European.

Nothing prevents them from reserving interesting samples for full sequencing.

Nothing prevents the owner of any bathroom you use from doing the same thing except, in both cases, the danger of being found out.

You would never get a clean sample that way.

Reply on news.ycombinator.com