The intent of this site to assist people with health issues that are, or could be, microbiome connected. There are MANY conditions known to have the severity being a function of the microbiome dysfunction, including Autism, Alzheimer’s, Anxiety and Depression. See this list of studies from the US National Library of Medicine. Individual symptoms like brain fog, anxiety and depression have strong statistical association to the microbiome. A few of them are listed here.
The base rule of the site is to avoid speculation, keep to facts from published studies and to facts from statistical analysis(with the source data available for those wish to replicate the results). Internet hearsay is avoid like the plague it is.
People with mast cell issues typical treatment approach is:
Reduce foods high in histamine
Avoid probiotics that produces histamine
Use DAO supplements.
There is a fourth leg that should be added to this three legged stool. Probiotics (and other substances) that impacts DAO levels. Some could argue that they impact DAO because they produce histamines. That may be true in some rare cases, but many of the items listed below are not known to produce histamine.
The production and activity of DAO can be regulated by various factors, including hormones, inflammatory mediators, and the presence of substrates like histamin. Difference probiotics produces different enzymes which may inhibit or assist with the production of DAO.
The probiotics, impact and study used are listed below:
Reviewing the studies is recommended. In some cases DAO levels were raised due to infection, i.e. Escherichia coli K88; and the probiotic may be inhibiting the increase due to the infection though reducing the infection and not actually impacting DAO levels.
“nicotinamide partially inhibits the in vitro release of histamine ” [1963]
“Release of PGD2 [due to niacin] was not accompanied by a release of histamine which was assessed by quantification of plasma levels of the histamine metabolite,” [1989]
CONCLUSION: Niacin by itself does not cause histamine release (except with excessive use of one form of niacin)
“Moreover, it[Niacin] led to a significant rise in DAO levels on day 30 (p < 0.05). Niacin supplementation significantly reduced the LPS levels on day 30 (p < 0.05).” [1992]
What is needed are control studies measuring DAO levels in humans with different probiotics. The above illustrates that probiotics are likely to make good or bad differences.
1976-2012 – Sustained 7 concussions playing contact sports, minor short term symptoms and fully recovered. Fit, active, and employed as adult. Satisfying family/social life. Higher intensity physical training 5-6 days per week outside or in gym. No drugs/alcohol since 2003, no medical issues or medications.
2012 – Post Concussion Syndrome – Fell and hit head skiing, diagnosed with concussion, symptoms did not resolve. No detectable damage on MRI.
Symptoms
Fatigue, brain fog, intense and unpredictable nerve pain in head
Cognitive and visual processing issues
Sensitivity to light, loud noise, busy environments
GI – constipation
Multiple food, chemical, and environmental sensitivities
Overall functionality reduced, still able to socialize, drive, complete tasks, problem solve, etc… sporadically and in short increments. Physical fitness mostly unaffected, continued to exercise, incorporated yoga/gyrotonic.
Divorced
Received disability benefits in 2019
chronic sinus issues that started after concussion and worsened with vaccine.
diagnosed w nasal mold colonization, nasal biofilms, and chronic staph aureus.
2021 – mRNA Vaccine Injury (2 shots Pfizer)
Symptoms:
Total exercise intolerance, sympathetic activation, anxiety and confusion
Exacerbation of pre-existing symptoms – severe fatigue, brain fog, food/chemical/medication/supplement sensitivities, constipation
Neuropathy – pain, tingling, numbness in legs and feet.
Minimal activity, short walks, occasional driving short distances, most time spent resting
the plasma level of Zonulin was significantly increased after post blast exposure, indicating that blast may contribute to the impairment of the gut barrier in the paracellular pathway “
Repetitive blast exposure resulted in both similar (e.g., increased IL-6), and disparate (e.g., IL-10 increase only in females) patterns of acute serum and brain cytokine as well as gut microbiome changes in female and male mice. Chart is below
increased Clostridium innocuum and Erysipelatoclostridium
reductions in Bacteroides and Clostridium Sensu Stricto
Reductions in Bacteroides have been associated with irritable bowel syndrome (IBS) development and identified after stroke [87]. Bacteroides are imperative for the maintenance of intestinal barrier integrity, with supplementation being associated with increased tight junction proteins [88]. Reductions in Clostridium sensu stricto have been associated with reduced butyrate production and Alzheimer’s disease [89,90].
” Findings from this trial support the feasibility, acceptability, and safety of supplementation with an anti-inflammatory/immunoregulatory probiotic, Lactobacillus reuteri DSM 17938, among Veterans with PPC and PTSD symptoms.” [2020]
COVID Shots
Explicit studies are rare. What we do have is a variety of studies between high and low responder which enumerates the difference of bacteria. We do not clearly know which goes in any specific direction. We do know the bacteria that are likely to change.
Increase Akkermansia muciniphila in some cases, but in general no change
The high response group were primarily characterized by a predominance of Enterococcus faecium, Prevotella bivia, Actinomyces massiliensis, Veillonella dispar, Veillonella_sp_T11011_6, Eubacterium_sp_CAG_38, Ruminococcus torques, Actinomyces odontolyticus, while Alistipes putredinis, Allisonella histaminiformans, Bacteroides clarus, Clostridium lavalense, Clostridium asparagiforme, Bacteroides eggerthii, Coprobacter fastidiosus, Sutterella parvirubra, and Blautia coccoides are more abundant in the low response group
We have two results to work from: Biomesight report and an OATS (Organic Acid) report.
Looking at Forecasted symptoms we have a high rate of pattern matches. This is hopeful because it implies we now have strongly suspected bacteria.
We also see a significant shift of bacteria that are atypically over represented.
The reader noticed the dominance of some unusual bacteria (species and the genus they belong to). These amount to be 28.4% of the microbiome. These two bacteria usually average 3.5% of the microbiome.
As a starting point, we will do [Just give me Suggestions include Symptoms]. Since we have some condition specific suggestions above, let us see where they rank.
This type of cross validation is nice to see — everything known to help TBI/concussion is in the recommended to take list. It builds confidence in the suggestions being generated.
The next thing is to see which of the bacteria shifts cited in the above literature. We go over the the Bacteria tree. We filter out those not usually reported by Biomesight and ended with just one (which illustrates the benefit of shot-gun tests).
reductions in Bacteroides: was at 32%ile
Other incidental measures of note:
Anti Inflammatory Ratio: 20% (so inflammation is likely)
Other Issues
Last, do we see Staphylococcus aureus in the sample? No.
“Some (poly)phenolics such as caffeic acid [found in Barley, Coffee], hydroxytyrosol [from Olive Oil}, resveratrol, curcumin, nordihydroguaiaretic acid (NDGA) [TOXIC ISSUES], and quercetin have been reported to reduce the formation of 5-LOX eicosanoids in vitro”
I should point out that something is very funky with the OATS report. Every value below is within their declared range but their graphics show it is out of range.
The result was 326 probiotics — specifically, probiotics that are known to consume Malate. We have no significant additions.
Action Plan
First item of note, above we have 100% cross validation on what the fuzzy logic expert system says should help with clinical studies of what helps given his particulars. This hints that other suggestions are far more likely to help than hurt.
The easy set of suggestions are quercetin, clostridium butyricum (probiotics),lactobacillus acidophilus, lactobacillus reuteri (probiotics) because they are double recommended: both clinical studies and the microbiome. Since with this history, getting any prescription drugs is unlikely, the consensus suggestion leads me to consider the following (filtered for availability etc):
Probiotics (no more than 2 at a time, and rotate to different ones every 2 weeks)
IMHO, This is the general problem with diet data — they contain vast baskets of substances: some good and some bad; with a wide variety of definitions. I prefer working substance by substance.
Retest
I would suggest a retest with Biomesight in 3 months (if the reader consents, I would be glad to do a follow up post). Remember these are not generic suggestions for anyone with a concussion but based on the individual’s microbiome with cross referencing to the literature to develop a clean consensus of what may help.
Ideally, the user will do the time consuming process of checking the suggestions against the literature (which I did above for illustration).
There is NO direct linkage using studies of symptoms to supplements by the engine. That is technically possible, but would require major funding to hire qualified people to enter the data.
Experience have found that 85-95% of suggestions that has studies for a conditions are in agreement.
Q: Do you use studies on my conditions to pick bacteria?
If there are sufficient studies, then yes. In your case there is not. Clinical studies on conditions often have contradictory results for a vast number of reasons: the lab and software used; the diet of the people in the studies; often low significance (often P < 0.05 is cited, with our lab specific analysis we typically use P < 0.001 as a criteria).
Q: The high Bilophila on my Biomesight got my attention. Is that something that MP.com does not identify specifically to address but rather it just corrects as part of the overall microbiome rebalancing?
MP does a holistic analysis — so things that may reduce Bilophila but also shifts others bacteria in the wrong direction may be eliminated.
“The typical MP matrix to solve is around 60 taxa (up to 430 in some cases) by 2092 possible modifiers – thus an array of some 12,000 to 800,000 elements to consider. The Monte Carlo method typically uses 5 runs resulting in 60,000 to 4,000,000 elements evaluated.” [blog]
Q: The prebiotic that shows up as #1 rec for me is Prefor Pro, but you have chicory. Is that due to availability?
The retailed probiotic selection is based on the species in the probiotics ignoring additives and relative amounts (often not declared). If there are issues with these additives, then just move down the list.
Q: Do you think using the condition specific bacteria shifts is a better approach than the bacteria identified by the AI as “likely to be causing my symptoms”?
Definitely, my observations from feedback is that targeting those bacteria do moderate or eliminate symptoms.
Postscript – and Reminder
I am not a licensed medical professional and there are strict laws where I live about “appearing to practice medicine”. I am safe when it is “academic models” and I keep to the language of science, especially statistics. I am not safe when the explanations have possible overtones of advising a patient instead of presenting data to be evaluated by a medical professional before implementing.
I cannot tell people what they should take or not take. I can inform people items that have better odds of improving their microbiome as a results on numeric calculations. I am a trained experienced statistician with appropriate degrees and professional memberships. All suggestions should be reviewed by your medical professional before starting.
The answers above describe my logic and thinking and is not intended to give advice to this person or any one. Always review with your knowledgeable medical professional.
This is part of a series of technical notes. If this interests you, you may wish to view others in the series.
A colleague wrote me:
I am currently researching companies who are using machine learning and LLM approaches for Microbiome analysis I wanted to know your opinion why this approach fails/ is less precise than your [MicrobiomePrescription MP] Fuzzy LogicExpert System with Monte Carlo method.
I have been doing a wide variety of Artificial Intelligence development professionally since 1988 for firms including Microsoft, Amazon, Verizon and Starbucks. I also have a reasonable science background including High School General Science teacher and College Chemistry and Physics instructor.
In the early days of Microbiome Prescription I tried a panacea of technics such as Random Forest, linear and non linear regression, supervised learning, etc. The results were less than acceptable. A friend, Richard Sprague, also ex-Microsoft, who worked as Citizen Scientist in Residence for uBiome for a while set up a series of meeting with the teams of the Allen Institute for AI [AI2]. The consensus working with those experts was that my direction of using the expert system model was far superior than what they could come up with.
How does Machine Learning and LLM work?
The core mechanism is pattern recognition of key words and phrases. This allows a numeric representation of the literature to be created, i.e. [subject #,verb #, object #]. When someone asks about subject, a set of equivalent subject # is obtained, a set of equivalent verb # is obtained and we just lookup the data. The number of records may be used to determine priority or most probable outcome.
This leads to the problem with many implementations, phrase recognition. To illustrate consider two articles, one mentions: Limosilactobacillus reuteri and the other mentions for Lactobacillus reuteri. A microbiologist knows that these are the same, but the typical ML or LLM does not. Do we end up with two collections of information? Since both Limosilactobacillus and Lactobacillus are also taxological units, do we get this species information incorrectly applied to the genus?
This is further complicated going to studies dealing with strains. Compare these two species: Escherichia coli O157:H7 and Escherichia coli Nissle 1917. One causes food poisoning, and the other is a probiotic shown to improves Crohn’s disease. Strains often have multiple identifiers and retail names. For example, Enterococcus faecium SF68 is also known as Enterococcus faecium NCIMB 10415 and sold retail as Bioflorin for humans and multiple brands in veterinary practice.
So the number 1 issue is correct identification. We also have some name collisions: Bacteroidetes is the name of both a class [Taxon 200643] and a phylum [Taxon 976] (which is now called Bacteroidota). A knowledgeable reader can reason out what is meant.
Taxonomy Hierarchy Inference
Information on the microbiome is sparse. A microbiome practitioner asked about what would reduce Lactobacillus balticus may discover that there is no literature on it. This practitioner would then infer that whatever reduces Lactobacillus would reduce it. This is an inference which MP does (remember it is an expert system mimicking the behavior of a human expert). MP takes it one step further by recognizing that is now classified as Limosilactobacillus balticus. Instead of items impacting Lactobacillus, it will use items impacting Limosilactobacillus. The suggestions are more probable to be correct.
What to address question
The various AI systems may well scrap ranges from studies to apply to a microbiome sample. Results are not consistent from lab to lab [the back story is this 2019 post: The taxonomy nightmare before Christmas…] . This means that reference ranges are more inconsistent as a consequence. Below are some range examples from commercial tests
The reference range dilemma is why MP generates suggestions based on reference ranges determine by multiple methods: Average +/- Std Dev, Box Plot Whiskers, ranges from specific sources (including the processing lab in some cases), and patent pending algorithms. Each reference range is determined from a large collection of sample processed by the same lab. The suggestions using each of these reference ranges are then aggregated into a “consensus” report (i.e. Monte Carlo method).
“We are using AI to get Suggestions” – marketing hype!
A few years ago, microbiome testing companies would attempt to get creditability by claiming their suggestions were created by registered dietitian. Today, “AI” is the replacement. If you ask about what AI methodology is being used, the size and scope of the data behind, etc. 99% of the time you will be given a “It’s proprietary! We cannot disclose it“.
They may be truthful that is coming from AI, for example, someone asked https://www.perplexity.ai/, “Which foods reduces Fusobacterium” (Example answer – Fusobacterium nucleatum is what is cited). They copy the answer into their database to show their customers. Thus it is true that the answer is coming from AI; but it is a one-dimensional blinkered answer that will often leaves their customer worse. This AI is ignorant that Fusobacterium prausnitzii [The bacteria formerly known as Faecalibacterium prausnitzii] belongs to it. This bacteria has lots of studies that are ignored by the AI! The AI appears ignorant that two studies report both quercetin, mastic gum (prebiotic) reduces it. It likely has the data but has misclassified it.
My observations of reviewing many sites is that suggestions are scoped to a single bacteria and ignores side-effects on other bacteria that would also be impacted. MP uses holistic algorithms [which is what would be expected when someone has done Ph.D. courses on Integer and Non-Integer Programming Optimization]. The typical MP matrix to solve is around 60 taxa (up to 430 in some cases) by 2092 possible modifiers – thus an array of some 12,000 to 800,000 elements to consider. The Monte Carlo method typically uses 5 runs resulting in 60,000 to 4,000,000 elements evaluated.
Examples of LLM gone bad
Most experts know that Mutaflor is Escherichia coli Nissle 1917 and is clearly names as such in publicly accessible papers on the National Library of Medicine. So this response is one of AI’s famous hallucinations; hallucinations are not possible from expert systems.
Bottom Line
Microbiome testing firms may correctly claim[in a legal sense] they are AI based. If they refuse to fully disclose the methodology being used (ideally on their site), then the safest assumption is that they got a summer intern to ask one of the LLM’s the questions and just copied the answers into the database. Without full disclosure, they simply cannot be trusted.
If you are considering using AI because some “hot shot evangelist or venture capitalist” is pushing for it; then — look at the above issues and insists on documentation on how each of these issues will be addressed. Until there is clear, understandable documentation on these issues, “The suitability of AI has not been shown for the proposed AI implementation” and stop wasting time and money!
MP uses a very old model of AI that requires manual data curation being feed to the expert system. This a hallmark of expert systems.
I wish to give a special thanks to Juan Pablo C., Assistant professor in Universidad Mayor, Chile and former Senior Bioinformatician for uBiome for bringing to my attention the article below. He was then kind enough to point me at appropriate data sources to allow me to implement this new measure.
Oral bacteria relative abundance in faeces increases due to gut microbiota depletion and is linked with patient outcomes
The detection of oral bacteria in faecal samples has been associated with inflammation and intestinal diseases. The increased relative abundance of oral bacteria in faeces has two competing explanations: either oral bacteria invade the gut ecosystem and expand (the ‘expansion’ hypothesis), or oral bacteria transit through the gut and their relative increase marks the depletion of other gut bacteria (the ‘marker’ hypothesis). … By distinguishing between the two hypotheses, our study guides the interpretation of microbiome compositional data and could potentially identify cases where therapies are needed to rebuild the resident microbiome rather than protect against invading oral bacteria.
I have implemented a new measure on Microbiome Prescription to estimate the amount of leakage. It is located in the Health Analysis section (a large and often slow loading page)
The report gives a percentile ranking based on the total for genus that are Oral Bacteria and then lists the specific genus and species present
As with other study specific pages, you can hand select bacteria of interest or click on the bacteria to get more information about the bacteria. Of special note is Prevotella being listed as there is evidence suggesting that it increases due to fungi in the environment – thus the gut microbiome depletion is likely why it can take up residency.
I will be looking to modify the suggestions algorithms to exclude oral bacteria and thus have the suggestions focused on gut microbiota depletion as a treatment option.
“For the Ombre/Thryve sample data, dated 12/31/24, I found the following:
Rickettsiales – order – 31%ile High
Rickettsiaceae – family – 34%ile High
Rickettsieae – tribe – 34%ile High
Parasutterella excrementihominis – species – 69%ile High
The reason these bacteria stand out to me are because my CFS journey, like many, began with a Lyme diagnosis 20 years ago. And as testing and research improved, Lyme became synonymous with “co-infections” like Rickettsia and Ehrlichia and a whole host of others in the “Lyme Soup”.”
Parasutterella excrementihominis has been present in all samples, varying from 61-89%ile. For Rickettsiaceae, we see in much older samples
2022-04-11 at 6%ile
2022-11-01 at 18%ile
All bacteria wax and wane over time. Personally, I would not be concerned unless it is constantly above 75%ile for pathogenic. It is is over, then take your concerns to your MD for appropriate testing to confirm..
For this person we have a lot of symptom forecasts[new algorithm] matching reported symptoms
Analysis
Ombre Lab Processing
We are looking at the Ombre Lab Report data below. These are a lot of swings over the last 4 sample. Looking at the prior (likely the worse)
Eubiosis was the worse
Bacteria Under 10%ile was the highest
Outside Kaltoft-Moldrup was the highest
Chao1 Index was the highest
A lot of other measures had little or no change (i.e. JasonH, Microba Co-Biome,,Nirvana/CosmosId etc) suggesting low sensitivity to detect shifts. Some hits at a continuous shift (Simpson Diversity Index.
The wide variety of lab read quality makes reliable comparisons difficult. The Lab Read Quality bounces around, and with that, other values may echo these shifts (i.e. up to 20% shifts for some measures). A low read quality means less bacteria are reported, for example, when it was low, the Outside Kaltoft-Moldrup has low, when it was high, the value became high.
Another way to view it is this: If 10% are out of range and 660 are reported then we have 66. If we have 940 in another report then we would expect 94. This could be misread as a 94/66 or a 43% increase in out of range bacteria. Technically, it is more complicated but that should explain the problem.
Criteria
12/31/2023
10/2/2023
7/28/2023
4/18/2023
Lab Read Quality
5.6
12.6
8
4.9
Eubiosis
53.4
8.2
38.8
83.1
Rickettsiaceae
34%ile
0
0
0
Outside Range from GanzImmun Diagostics
17
17
14
14
Outside Range from JasonH
5
5
5
5
Outside Range from Lab Teletest
32
32
25
25
Outside Range from Medivere
17
17
17
17
Outside Range from Metagenomics
8
8
7
7
Outside Range from Microba Co-Biome
5
5
6
6
Outside Range from MyBioma
16
16
9
9
Outside Range from Nirvana/CosmosId
24
24
26
26
Outside Range from Thorne (20/80%ile)
310
310
210
210
Outside Range from XenoGene
51
51
41
41
Outside Lab Range (+/- 1.96SD)
36
20
7
19
Outside Box-Plot-Whiskers
114
97
71
69
Outside Kaltoft-Moldrup
164
239
141
79
Bacteria Reported By Lab
765
941
762
663
Bacteria Over 90%ile
101
82
42
58
Bacteria Under 10%ile
60
302
102
28
Shannon Diversity Index
2.953
3.099
2.987
2.62
Simpson Diversity Index
0.043
0.053
0.079
0.098
Chao1 Index
23041
38183
26135
15707
Pathogens
28
25
23
15
Condition Est. Over 90%ile
1
1
1
1
Top 10 forecast symptoms matching
9
9
10
n/a
Rickettsiaceae was seen in earlier samples:
2022-04-11 at 6%ile
2022-11-01 at 18%ile
Biomesight Reporting
One sample could not be successfully processed thru Biomesight. Biomesight indicated that there are ongoing technical issues with their processing of Thryve FastQ files.
Criteria
12/12/2023
7/28/2023
4/18/2023
Lab Read Quality
5.6
8
4.9
Eubiosis
42.8
80.8
74.3
Outside Range from GanzImmun Diagostics
13
14
14
Outside Range from JasonH
3
5
5
Outside Range from Lab Teletest
16
23
23
Outside Range from Medivere
12
13
13
Outside Range from Metagenomics
5
8
8
Outside Range from Microba Co-Biome
1
2
2
Outside Range from MyBioma
8
4
4
Outside Range from Nirvana/CosmosId
11
23
23
Outside Range from Thorne (20/80%ile)
144
250
250
Outside Range from XenoGene
20
23
23
Outside Lab Range (+/- 1.96SD)
38
22
25
Outside Box-Plot-Whiskers
149
97
81
Outside Kaltoft-Moldrup
99
123
60
Bacteria Reported By Lab
664
696
618
Bacteria Over 90%ile
122
93
61
Bacteria Under 10%ile
38
46
9
Shannon Diversity Index
1.565
1.242
1.507
Simpson Diversity Index
0.037
0.057
0.078
Chao1 Index
17542
14092
10705
Shannon Diversity Percentile
41.2
10.8
32.6
Simpson Diversity Percentile
32.4
55.9
72
Chao1 Percentile
88.4
75.6
53.8
Lab: BiomeSight
Pathogens
25
23
19
Condition Est. Over 90%ile
1
1
1
Top 10 forecast symptoms matching
7
n/a
9
Bottom line, the variability of 16s Lab Quality leaves too much fogginess for interpretations 🙁
Revised Symptom Forecasting
This algorithm is similar to the Eubiosis algorithm. We compute the expected number of matches to bacteria shifts associated with the symptoms. The expected theoretical threshold by randomness is 16%. A higher number indicate increased odds, a lower number decreased odds. This is based on the existing annotated samples uploaded. It is not definitive and often there can be multiple subsets of bacteria associated with a symptom. The match is on too much or too few of a collection of bacteria
The checkmarks are the entered symptoms, the list are the predictions from most likely to lesser.
As seen in another post over multiple samples, we have a good accuracy in predicting symptoms from the microbiome.
Going Forward
Again, using Just give me suggestions include Symptoms is how we are going to proceed. And then add in the two Special Studies. This results in 7 packages of suggestions for each lab. Since we have two different interpretations (Biomesight and Thryve) of the FastQ files. We then do an Uber Consensus merging 14 packages of suggestions!
And the merge (on [Multiple Samples] tab)
Thresholds: High is 383 thus 190 or higher, Low is -462 this -231 or lower
First we are going to look on what ALL 14 sets of suggestions agreed upon
Remember to check suggested dosages here. A common issue is taking token amounts which are unable to effect changes.
Postscript – and Reminder
I am not a licensed medical professional and there are strict laws where I live about “appearing to practice medicine”. I am safe when it is “academic models” and I keep to the language of science, especially statistics. I am not safe when the explanations have possible overtones of advising a patient instead of presenting data to be evaluated by a medical professional before implementing.
I cannot tell people what they should take or not take. I can inform people items that have better odds of improving their microbiome as a result on numeric calculations. I am a trained experienced statistician with appropriate degrees and professional memberships. All suggestions should be reviewed by your medical professional before starting.
The answers above describe my logic and thinking and is not intended to give advice to this person or anyone. Always review with your knowledgeable medical professional.
This person had been doing duel interpretation of their FastQ files (Thyrve and BiomeSight). The results are mixed. A common problem comparing results is not there, the Lab Quality is similar for most of the samples,
Criteria
2/2/2024
3/23/2023
10/29/2022
8/22/2022
7/21/2022
4/30/2022
Lab Read Quality
4.9
8.7
5.1
5
4.9
4.8
Eubiosis
100% 🙂
36%
50%
66%
87%
74%
Outside Range from GanzImmun Diagostics
13
13
18
18
16
16
Outside Range from JasonH
6
6
6
6
7
7
Outside Range from Lab Teletest
26
26
27
27
29
29
Outside Range from Medivere
17
17
17
17
17
17
Outside Range from Metagenomics
9
9
8
8
9
9
Outside Range from Microba Co-Biome
9
9
9
9
11
11
Outside Range from MyBioma
8
8
8
8
9
9
Outside Range from Nirvana/CosmosId
19 🙂
19
22
22
24
24
Outside Range from Thorne (20/80%ile)
217 🙂
217
300
300
216
216
Outside Range from XenoGene
48
48
53
53
45
45
Outside Lab Range (+/- 1.96SD)
26
4
6
16
8
4
Outside Box-Plot-Whiskers
104
34
36
120
73
44
Outside Kaltoft-Moldrup
126
161
85
125
102
70
Bacteria Reported By Lab
694
632
604
843
752
575
Bacteria Over 90%ile
65 🙁
28
30
87
47
37
Bacteria Under 10%ile
76
140
28
61
51
25
Shannon Diversity Index
3.107
3.028
2.987
3.227
3.126
3.121
Simpson Diversity Index
0.044
0.085
0.076
0.045
0.088
0.075
Chao1 Index
16886
13972
13231
27357
20690
9772
Pathogens
29
30
25
34
35
28
Forecasted
4
3
n/a
n/a
4
4
We have a few matches with the latest symptom forecast. Remember this is pattern matching over symptoms that likely have dozen of subsets.
Going Forward
We are going to do the [Just give me suggestions include Symptoms] . This gives us 5 packages of suggestions.
Looking only at what is suggested by all 5 sets of suggestions
Highest value was 280, thus a threshold of 140, lowest value was -220, thus a threshold of -110. What was VERY DIFFERENT then most samples, there were not a swarm of prescription items hogging the top. Also the range of priorities was much less than most samples examined. This implies less severe dysbiosis (or less that we can do, i.e. suggestions).
I am not a licensed medical professional and there are strict laws where I live about “appearing to practice medicine”. I am safe when it is “academic models” and I keep to the language of science, especially statistics. I am not safe when the explanations have possible overtones of advising a patient instead of presenting data to be evaluated by a medical professional before implementing.
I cannot tell people what they should take or not take. I can inform people items that have better odds of improving their microbiome as a results on numeric calculations. I am a trained experienced statistician with appropriate degrees and professional memberships. All suggestions should be reviewed by your medical professional before starting.
The answers above describe my logic and thinking and is not intended to give advice to this person or any one. Always review with your knowledgeable medical professional.
I suspected that this may be connected to Mast Cell Activation Syndrome (MCAS), a condition that many people struggle with. I thought it would be good to do a walk-through of how I currently research things that are of interest but not explicit in the literature.
My first step is to see if it is worth while exploration. At present, I use Perplexity because it provides references for what it answers. You need to check them, one reference that I got on a different question came from [sarcasm]”the definitive medical resource: Amazon Product Information“.
The [answer] was: “Circulating inflammatory cytokines can have a significant impact on Mast Cell Activation Syndrome (MCAS)…“. If you click on [answer] you will see the full response.
In terms of compounds we have items which we lack microbiome data for, BUT as you will see later, we have herb data for!
CCL23 increases risk [Chemokine (C-C motif) ligand 23 (CCL23) is a small cytokine belonging to the CC chemokine family that is also known as Macrophage inflammatory protein 3 (MIP-3) Wikipedia ]
SIRT2, ADA and Caspase 8 decreases risk
We now have a bunch of scents to follow. Some may bear fruit.
I inspected these lists (which are at the genus level), and then look for matches. There were none 🙁
I went over to a sample from a MCAS person done with shotgun (Thorne) and a parallel Biomesight and Ombre test done at the same time. For the positively associated there were only one, reported by Ombre at 39%ile). For negatively:
Escherichia was 50%ile (33%ile from Biomesight, 59%ile from Ombre)
So the person appears to be deficient in bacteria that inhibits things. Incidentally, this person takes Mutaflor and SymbioFlor-2 (Escherichia probiotics) and have reported that it seems to moderate histamine issues.
Checking KEGG for scents:
SIRT2 – we see NAD-dependent protein deacetylase sirtuin-2 isoform X1, this corresponds to this person feeling better from taking (flushing) niacin. Perplexity clarifies it with “Niacin (also known as vitamin B3) is a precursor for the coenzyme nicotinamide adenine dinucleotide (NAD).” and five references!
ADA – adenosine deaminase connected to [EC:3.5.4.4] which we can seen on the site, this person BiomeSight was at 98%ile by Biomesight and 43%ile by Ombre, so situation is fine
For this person, the study and experience has reasonable agreement with this person’s positive response to Escherichia probiotics and (flushing) niacin appearing to be explained.
Phrase 2
This starts by using Perplexity to find branches of the above to explore.
What food or herbs increases Eubacterium ruminantium? Turmeric!!! BUT THIS WAS A HALLUCINATION! The AI deemed [TRF] to be turmeric residue fiber, but in the study it was Time-restricted feeding (TRF). — this illustrates the need to check the source every time!!
Now, if we go to Microbiome Prescription for these, example Sellimonas we have many dozens of substances with some being explicit. Clearly, Perplexity is inferior for bacteria information.
Next, we go and look at things not on microbiome prescription.
“Niacin-rich foods: Niacin (vitamin B3) is a precursor for NAD biosynthesis in the body. Therefore, consuming foods rich in niacin can provide the raw material for NAD production.”
“Tryptophan-rich foods: The amino acid tryptophan can also be converted to NAD in the body, albeit at a lower efficiency than niacin.2 Foods rich in tryptophan include:Turkey, Oats, Bananas, Milk, Yogurt” The quality of references were poor (very poor), so I did a PubMed search and found:
“Underlying these benefits [of moderate excise] were transcriptional changes in enzymes driving the conversion of tryptophan to NAD(+), this leading to increased hepatic NAD(+) and elevated activity of the NAD(+)-” [2021]
So tryptophan rich food plus moderate exercise would be an alternative for people who cannot use (flushing) niacin.
I decided to dive into PubMed for Caspase 8 with herbs and got 91 results. Some interesting ones are below. WARNING: Some of the herbs below are prohibited to be sold in the US because of risks. They are listed here for illustration only:
“The expressions of caspase–8 and caspase-9 were significantly elevated by the Kerra extract ” [2023]
“Furthermore, it[ leaf extract of Acorus calamus] also increased expressions of caspase-3, caspase-9, caspase-8,” [2023] FYI, available for sale. [WARNING from Web MD]
“Tribulus terrestris (TT), a herb belonging to Zygophyllaceae family …caspase 8 genes were also upregulated” [2019] FYI, it can be purchased as a powder in bulk
“extracted from the Chinese herb Tripterygium wilfordii Hook…was associated with activation of both caspase-3 and caspase-8” [2006]
“isolated from Solanum incanum herb … activation of caspase-8” [2004]
Of special interest is that most of these studies were in the context of cancer with caspase-8 inhibiting the cancer. MCAS and cancer are associated: ” Our data support the view that mast cells may promote development of certain malignant tumors. These findings indicate a need for increased surveillance of certain types of cancer in MCAS patients irrespective of its individual clinical presentation.” [2017]
I did a second dive into CCL23 (and checked it alternative names too!)
“CCL23 is produced predominantly by mast cells in systemic mastocytosis, and CCL23 plasma levels are associated with disease severity, correlating positively with established markers of disease burden, thus suggesting that CCL23 is a specific SM biomarker.” [2023]
This means that reducing TNFα should reduce CCL23. [perplexity gives a list with reference including: Trikatu, Allium sativum, Capsaicin, Carvacrol,Thymol. ]
Usually I work on microbiome modifications. The reason is simple — it is pretty time consuming to maintain the database! Above, we see a secondary approach — targeting specific enzymes and compounds using PubMed. I would love to get funding to put this extra information into an equivalent (and connected) expert system.
Bottom Line
The above illustrates how AI (Large Language Models – ChatGPT and Perplexity) can assist in researching — but beware of hallucinations! and check the sources of the information.
What is interesting is that we cascaded from Circulating inflammatory cytokines to MCAS to specific compound and then to herbs. The herbs are often atypical herbs that should be researched for safety and any other known issues. Many appear to be used traditionally in eastern medicine. This approach gives new options to consider when conventional options fail. We do not know the microbiome impact of these herbs — but we do know the impact on specific compounds and enzymes of interest. I would love to build a fuzzy logic expert system in this area so that contraindications are visible and the results are sounder; unfortunately that takes skilled people and time.
As always, before starting any herbs, spices and supplements, you should review your plan with a licensed medical professional.
There are many approaches that can be taken. With the same data, different methods can reap a huge variation of harvest (i.e. the number of statistically significant relationships found).
I will use the Pearson’s Chi2 to determine significance and a simple “does the presence of bacteria A results in a shift of bacteria B”. By a shift, I mean either an increase/decrease of the number of bacteria above a threshold or below count. The threshold for bacteria count varies from bacteria to bacteria, and lab to lab.
To illustrate:
Bacteria Present
Bacteria Impacted
Low Zone
High Zone
Lactobacillus
Rickettsieae
528
185
The naive “expected number” is (528+185)/2=356.5, The actual expected number is lower 198 for both zones, yielding a chi2 of 647! The middle values have shifted lower. The conclusion that having Lactobacillus reduces the risk of Rickettsieae appears confirmed in publications such as Use of Lactobacillus to prevent infection by pathogenic bacteria [2002].
Looking in the opposite direction, we have a reduced Chi2 of just 29 with the apparent intrepretaion being that the prescience of Rickettsieae reduces Lactobacillus slightly (‘Odds ratio’ of 1.35, versus 2.66 going the other direction).
Bacteria Present
Bacteria Impacted
Low Zone
High Zone
Rickettsieae
Lactobacillus
268
202
IMHO, this approach yields more significant findings than looking at the differences of averages using standard deviations. Some people will attempt to find a linear regression between the counts of Lactobacillus and Rickettsieae. An example using the same data is below. I will leave it to the reader to reconcile.
A key philosophical question is whether we need to use all data, or only data that is significant. My resolution of this question is that the high and low levels are what is significant and the middle data is effectively just noise. Determining the cutoff points with backing mathematics/statistics is essential.
I have put interactions between bacteria of the same rank into Look up a bacteria taxa web site. Just search for a bacteria of interest and click the link.
You will see two charts, impacts and impacted by. For our example bacteria, we see that three other genus increases it (and looking at the names, no surprise!)
The size of the circles reflect the relative average count scaled
The width of the lines reflect the relative chi2 (significance), thicker lines implies more impact.
Green indicates increases/feed
Red indicates decreases/reduces
And a ton of bacteria that is reduced by its presence.
Applying to a Sample
Many of the bacteria above are rare. With an explicit sample, we filter to what is shown in the sample. We display them with their percentile ranking (i.e. a good indicator of relative amount)
Data Availability
The data (over 400K statistically significant impacts) will be available at MicrobiomePrescription Citizen Science for anyone wishing to compare against their own data. The data used was from Biomesight, a 16s provider that ships worldwide.
For me it’s still LongCovid > ME/CFS (thanks to SarsCov2) and unfortunately, in February, I had to take a 14-day course of antibiotics (amoxicillin 100 mg 3 times a day) because of Helicobacter Pylori
and also Pantoprazole 40 mg twice daily), which my micorobiome certainly didn’t like.
My PEM is less frequent and not as terrible as it used to be. My baseline has also improved, but I’m still at Bell 40 and pacing a lot. You already know the rest of my story.
Unfortunately, my daughter Carlotta looks as if she is slowly moving from the LongCovid control group to the LongCovid affected group. Which of course makes me particularly worried about my ME/CFS. She is also quite hypermobile.
Regarding her history, it should be mentioned that she developed bad migraines when she was around 5 years old (now 17). Which was triggered by certain foods. On a test it showed at 60 out of 80
Food intolerance. This could be almost completely remedied through a strict diet. The migraines were now rare, but have become more frequent recently. (maybe also interesting, there is a suspicion of Asperger’s Syndrome / Autism Level 1)
Mother
It seems that the microbiome has gone downhill over the year.
US National Library of Medicine Pattern Matching
2023: multiple chemical sensitivity [MCS], SIBO, Graves’ disease, Acne, hypertension
2024: hypertension, Menopause
Symptoms
2023: Not entered
2024: 101 Symptoms …
Dr. Jason Hawrelak Criteria: 66%ile
Daughter
The same pattern of the microbiome going downhill over the year.is seen here.
2023: Nothing entered (entering long afterwards is discouraged for the sake of accuracy)
2024: Neurocognitive: Can only focus on one thing at a time, Neurological: Joint hypermobility, Need to nap during each day, Impaired Memory & concentration, Onset: Gradual, Headaches, Migraine, Viral infections with prolonged recovery periods, Joint: Tenderness, Official Diagnosis: COVID19 (Fully Recovered), Acne, Difficulty falling asleep, Easily irritated, Tinnitus (ringing in ear)
Dr. Jason Hawrelak Criteria: 13%ile (i.e. bad)
Going Forward
The daughter’s Bifidobacterium was at 5%ile (extremely low with few species) and the mother’s at 29%ile with many species at low levels. This leads directly to my next observation.
My personal experience post-COVID was that a mixture of Bifidobacterium probiotics cleared a lot of symptoms in less than 2 weeks. I tried that based on the first published study below, with support from other studies:
A synbiotic preparation (SIM01) for post-acute COVID-19 syndrome in Hong Kong (RECOVERY): a randomised, double-blind, placebo-controlled trial [2023]
“Overall, Bifidobacterium was associated with both protective effects and reduced abundance in relation to the disease. The genus has been found to be abundant in some cases and linked to disease severity. The studies evaluating the use of Bifidobacterium as probiotics have demonstrated the potential of this genus in reducing symptoms, improving pulmonary function, reducing inflammatory markers, alleviating gastrointestinal symptoms, and even contributing to better control of mortality. In summary,” [2023]
“Growing evidence demonstrate that gut microbiota alteration is associated with COVID-19 progress and severity, and post-COVID-19 syndrome, characterized by decrease of anti-inflammatory bacteria like Bifidobacterium” [2023]
“Specifically, it suggests an association of anti-inflammatory bacteria, including Bifidobacteria species and Eubacterium rectale, with lower severity, and pro-inflammatory bacteria such as Prevotella copri with higher severity. ” [2022]
“Although the mortality rate was 5% in the [Bifidobacterium] probiotic group, it was 25% in the non-probiotic group. ” [2021]
“positive patients overall had lower relative abundances of Bifidobacterium ” [2022]
Thus, I was interested in what the KEGG Probiotics Suggestions came up with:
These KEGG suggestions appear to agree with the literature. See Explanation of the methodology if you are interested in the mechanics of these suggestions. I also looked at the revised supplements from KEGG (just done). Remember, using KEGG is not trying to fix individual bacteria, rather to make sure all of the nutrients needed in the “microbiome soil” are there in the hope of producing a bountiful healthy crop.
For the mother
NADH (due to low 3-Oxoadipyl-CoA), alternatively, regular niacin
Since she has acne, and the first two are often prescribed for acne… it may not be that hard to these prescribed.
What I found very interesting is the great similarity between suggestions of the mother and the child. Same DNA, similar diet, and likely similar time since COVID. I will leave them to review the avoid list. I will point out that the other B-Vitamins are on the avoid list. I should also point out that NADH and Niacin are closely related so we have agreement between the KEGG data and our usual expert system. Vitamin K is not often on our expert system list (little data to work from).
Key Take Away:
Depending on finances, retest after being on it for 6-12 weeks. Correcting the microbiome is usually a long list of course corrections.
Rotate and change Probiotics (maximum time on any one should be two weeks).
At least 10 BCFU for each probiotic species
Have Barley Porridge each morning with some butter
There was no strong converge in diet style (diet is very subjective with most studies and usually problematic to interpret). The following are specific items you should consider
I am not a licensed medical professional and there are strict laws where I live about “appearing to practice medicine”. I am safe when it is “academic models” and I keep to the language of science, especially statistics. I am not safe when the explanations have possible overtones of advising a patient instead of presenting data to be evaluated by a medical professional before implementing.
I cannot tell people what they should take or not take. I can inform people items that have better odds of improving their microbiome as a results on numeric calculations. I am a trained experienced statistician with appropriate degrees and professional memberships. All suggestions should be reviewed by your medical professional before starting.
The answers above describe my logic and thinking and is not intended to give advice to this person or any one. Always review with your knowledgeable medical professional.
There are two approaches to identifying bacteria associated with a group of symptoms:
UNION — you just join the bacteria associated with each symptom into a single list. This is often done when there is not sufficient data. It’s simple to do.
INTERSECTION — this identifies all people with the same combination of symptoms and then identify what is associated. This requires statistical computations to be done each time.
The video below is a quick walkthrough. What is interesting to note is that the number of significant bacteria can increase as more symptoms are added. Why? because you are filtering out noise from the bacteria.
You can also have bacteria appearing that were not in the prior list by adding one more symptom. Example below.
Bottom Line
With a large enough sample and enough characteristics recorded, you can drill down into a lot more data using the appropriate statistical techniques.
Recent Comments