23andme Patient Data Analysis

Dear members,

We are working to conduct a Genome-wide association study(GWAS) across patient groups based on the SNP data provided with the DTC genetic testing service 23andme.

We will be using high powered software and computing resources to perform a research project for the community. We are hoping to identify associated SNPs, and possibly observe similarities between those suffering the respective conditions at the genomic level. It is hoped that this information will help us and our scientific contacts better understand possible underlying genetic susceptibilities that could contribute to development of these syndromes. This is in contrast to the majority of people therapeutically using these substances who do not suffer persistent and devastating health problems following cessation. If statistical analysis of symptomatic vs control data sets reveals interesting findings, we will endeavour to work with one or more scientists to help us deepen our investigation and publish the results in a peer reviewed journal.

To achieve this, we need your help. If you already have taken a 23andMe test in the past, please click here for instructions on how to obtain your raw data. If you have not taken a 23andme test and would like to contribute your personal data to our project, please consider ordering a 23andme at-home saliva sample kit from their website. The cheaper ancestry option is sufficient as we require the raw data file, which is identical with either product offered.

Once you have downloaded and extracted the raw text file, please rename the file before submitting including your forum username and substance taken in the following way:


Examples would be as follows:


This will help us track which users have submitted data and their respective conditions.

Once your file is named appropriately, please submit your raw data file (once) via the following link. By submitting your data, you consent to the use of the data for the aforementioned purposes.

Thank you for your participation

– propeciahelp staff


Very nice.
How about including the number of datasets in the survey header with second color?

1 Like

If I pass this test, put something like my checkmark on it. I want to see who’s really responsible and who’s just talking and doing nothing. I really liked the idea of checking the box with the online survey.

1 Like

This is great!

I submitted my 23andme results a while back, though. Should I consider this to be a new initiative and submit again?


Thank you for your support. To clarify, this is the same initiative, we have simply split this information into its own thread. As this is part of a significant ramping-up of activities, we are now providing all the projects in one place with the new “Community Projects” category to prevent the important initiatives getting lost in the noise of the forum.



1 Like

If you want more 23 and me data you need another seperate marker like the blue tick for the survey and potentially another counter in the header as well. Otherwise most people won’t be aware of the project or will be too lazy to do it. Heck, even I haven’t done it, since I know you won’t get enough data anyway if you fail to aggressively promote it. I will do this but only if you promote it properly.

Does this gene test cover the whole gene sequence?

Unfortunately not, it is only a small percentage of the SNPs in the genome. However, considering the money spent by patients on repeated tests and treatments with no plausible benefit, we consider the low aggregate cost to be worth it for the potential benefit. This has the potential - even if it’s only a chance - to reveal information that could move us forward. It is worth taking that into account when considering if members can personally afford to take part.



Thanks @axolotl
Don’t you think that this could be missleading as we don’t even know which genes are affected, and what if the affected genes are not covered by this test? This doesn’t mean we should not do it, the only thing is, it could be tricky…


It’s a statistical frequency exercise vs control data, so no, I doubt there’s much chance for it to be misleading. It is possible that polymorphisms relevant to development of the condition are not in the selection. If that’s the case, then we just wont get a finding of significance. The cost is so low in aggregate it’s something we consider very worthwhile, and a project which offers the chance of a very important finding. Given the popularity of this test many members had it already. It isn’t the only iron we have in the fire - just an opportunity for patients who would like to help. We ideally need a fair few dozen more to have a data set of practical use, as GWAS is usually conducted with many more data sets than we have.


1 Like

I see there are 2 tests, one is 79 quid, the other is 149, probably the second one is more comprehensive I guess…?

The more expensive test gives you additional “reports” based on your data. We just need the raw data. So for our purposes the cheaper test is sufficient.