Genetic ancestry visualization app deployed on Heroku

I’d like to share an app that allows users to visualize their personal genome data (23andMe, etc.) in the context of the 1000 Genomes Project samples. The app projects user’s genetic data into a three-dimensional feature-space and predicts users’ genetic ancestry using a k-nearest neighbors classifier.

It would be great to hear ideas for improvements from the Streamlit community.
Thank you!

check out the app:

and the source code:


Very cool! Thanks for sharing @Kevin_Arvai!

Three very small usability ideas:

  1. I think I know what AFR, AMR, EAS, EUR, SAS are abbreviations for, but it would be nice to spell that out. Same on population. You could make those explanations hidden behind a checkbox. My general rule of thumb is to write out whatever someone might need to know - but if it’s supporting information then I hide it behind a checkbox so that it’s there if they need it but otherwise there’s not too much text on the page.

  2. I’d consider putting the toggle for super population vs population right next to the main graph. It’s a tough call to decide when to put something in the main app flow versus side bar, but if it’s something mainly impacting a visualization, it’s clearer to the user if it’s next to the visualization.

  3. As someone who has not done their own Genome, I’d still be super curious to see how it works with a sample data set. It would be interesting if you have one that I could choose as an example just to see how the app works.


Thank you, Amanda! Each of these are excellent suggestions to improve the app.
I’ve created issues in the GitHub repo.

I’ll post a reply when I finish them up :smile:

1 Like