This item has been on my backburner for a while and today I came up with an efficient algorithm to compute a short most probable symptoms. The results on my own samples were actually quite accurate.

The process is simple: Apply analysis (AI techniques) to the uploaded microbiome samples annotated with symptoms. We then flip the results to predict from an individual sample using these results. A lot of computational magic going in and out.

I have MTHFR issues (reported by conventional testing), this is my age range, cold hands and feet, I have been well aware since the 1970’s of eye focus issues (from eye specialist). The disorientation matches up with my being a high functioning autism spectrum person. The last item makes it a slam dunk….

I checked another person who I know the medical history. The person has TMJ, and just about every symptom listed!

Where is this new feature

One thing that I realized that it predicted symptoms that I would NOT associate with my current state– but actually matches the long term medical issues. For example, eye focus issues — that goes back 40 years and which is not on my radar as an active symptom.

I suspect “mouth sores” which showed up on other issues — may include gum disease being active.

Addendum

A reader try it and posted on facebook.

As a result, I added a link to the page to also give extended prediction with an indicator of the prediction confidence (score).