Existing US National Library of Medicine studies are insufficient often because they report a simplistic target group is higher or lower than the controls. The latest refactor of Symptoms (via Citizen Science) actually detect what I term “middle peaks”. Middle peaks can actually be adjusted to move someone away from a symptom’s clustering of bacteria range.

.

We are able to detect clustering of value connected to symptoms. These values may NOT be abnormal for everyone. When we compute the statistical values for those with the symptoms, we find that this group of values is abnormal.

The adjustment process is the same as for the extreme values — we want to shift the values towards the middle. If we move the values to the other side, that is actually good for improving the symptoms.

People Have Multiple Symptoms

In the past, my advice has been simple — figure out which is most important and address those. With the symptom-association refactor we can deal with multiple symptoms to a reasonable extent. I will be using actual data for a person with the following main symptoms

• GERD
• Crohn’s Disease
• Mast Cell Issue

Step #1 Examines the conditions and pick your options

After logging on (Important), we go to Symptoms with Bacteria Relationships and see what we have for each of the above

Step #2 Verify that you sample is a reasonable fit

The next step is simple, for each of the above (highest samples count) we click on the link. We end up with 3 different pages, Note (in blue) that the number of bacteria is less than above? Why? because different labs report on different bacteria. We, by design, ignore any bacteria that was not reported in the sample (if you disagree, all of the needed data is available on the citizen science site for you to create your own rules and analysis with)

Step #3 Get Suggestions for Each Reasonable Fit

All of the items above were good fits. So for each we click [Create Other Samples Profile for Selected]. Strong and Very Strong are automatically selected. You can adjust the selection with the checkboxes if you wish. On the next page, just click thru (after adjusting on any desired items) on the [Show Suggestions] button

You can look at each suggestion, but we have added a Consensus Report to make combining items easier. If you return to the 16s Sample Page you will see a new button appeared indicating that 3 sets of suggestions has been recorded.

Step #4 View Consensus Report

Click on the button and you will see a drop down, select View Consensus

The Report is in 6 sections (following the pattern for Probiotic)

• Absolute Takes — these are items with no known negative impact on any bacteria under consideration (for ALL of the suggestions sets). These are the safest items to add
• Probable Takes — these have some known negative impact, but the likely positive impact is very good
• Possible Takes — these have some known negative impact, but the likely positive impact is not as certain
• Absolute Avoid– these are items with no known positive impact on any bacteria under consideration (for ALL of the suggestions sets). These are not wise choices
• Probable Avoid– these have some known positive impact, but the likely negative impact is bad
• Possible Avoid — these have some known positive impact, but the likely negative impact is not good

The report lists the total number of items and allows you to restrict items to a specific level of impact confidence. The usual suggestions are all scaled so the maximum impact is 1.0 The numbers here are not scaled.

Note that we can get each table sorted by clicking on the column title.

Bottom Line and Caution

The first item is to not include symptoms that are not good fits (I am working on calculating the fit – coming soon). When I toss in other canned suggestions (Kaltoft-Moltup or Dr. Jason HawrelakRank Used:All Ranks) in, the results do not appear as good.

The second item is that it should be review by your medical professional. All of the items are based solely on the impact on the microbiome bacteria — often items may have adverse effect on medical conditions.

Again, this is done by AI and mathematical/statistical models — it is not based on clinical experience. It is not medical advice, it describes a methodology that should be discussed with your knowledgeable medical professional (where ever they are hiding).

REMEMBER: This is based on one individual microbiome and applies only to them. There are 215 bacteria for Mast Cells above. This person sample from a specific lab had 60 matches (about 30%) of which we used 56 to generate the suggestions. Another person with the same 3 items may have a totally different set of bacteria identified.