COPD and the gut microbiome

Chronic Obstructive Pulmonary Disease (COPD) is a group of diseases that make it difficult to breathe. It includes emphysema, chronic bronchitis and, in some cases, asthma [src].  A friend’s wife has it and I decided to see if there was microbiome shifts associated with it, in the naïve hope that adjusting those shifts may reduce the severity.

There are many studies on the lung microbiome [110+ at present] but modifying that microbiome does not have the usual tools available. The sole one that came to mind was Symbioflor-1, which has a study from 2001 showing a 40% drop in incidences, Influence of a bacterial immunostimulant (human Enterococcus faecalis Bacteria) on the recurrence of relapses in patients with chronic bronchitis This good Enterococcus faecalis probiotic appears to suppress the bad, drug resistanct, Enterococcus faecalis found in lung from earlier studies [1993], Symbioflor-1 is provided as a liquid that is effectively gargled with(“Keep Symbioflor 1 One minute in the mouth and gurgle in front of the swallow.“) This allows some of it to get into the throat and eventually the lung.

More Literature on Symbioflor-1 for those that are interested

A Comparative Transcriptomic Analysis of Human Placental Trophoblasts in Response to Pathogenic and Probiotic Enterococcus faecalis Interaction.
The Canadian journal of infectious diseases & medical microbiology = Journal canadien des maladies infectieuses et de la microbiologie medicale (Can J Infect Dis Med Microbiol ) Vol: 2021 Issue Pages: 6655414
Pub: 2021 Epub: 2021 Jan 28 Authors Tan Q , Zeng Z , Xu F , Wei H ,
Summary Html Article Publication
Influence of Catecholamines (Epinephrine/Norepinephrine) on Biofilm Formation and Adhesion in Pathogenic and Probiotic Strains of Enterococcus faecalis.
Frontiers in microbiology (Front Microbiol ) Vol: 11 Issue Pages: 1501
Pub: 2020 Epub: 2020 Jul 24 Authors Cambronel M , Nilly F , Mesguida O , Boukerb AM , Racine PJ , Baccouri O , Borrel V , Martel J , Fécamp F , Knowlton R , Zimmermann K , Domann E , Rodrigues S , Feuilloley M , Connil N ,
Summary Html Article Publication
In silico analyses of the genomes of three new bacteriocin-producing bacteria isolated from animal`s faeces.
Archives of microbiology (Arch Microbiol ) Vol: 203 Issue 1 Pages: 205-217
Pub: 2021 Jan Epub: 2020 Aug 17 Authors Eveno M , Belguesmia Y , Bazinet L , Gancel F , Fliss I , Drider D ,
Summary Publication Publication
Probiotic Enterococcus faecalis Symbioflor 1 ameliorates pathobiont-induced miscarriage through bacterial antagonism and Th1-Th2 modulation in pregnant mice.
Applied microbiology and biotechnology (Appl Microbiol Biotechnol ) Vol: 104 Issue 12 Pages: 5493-5504
Pub: 2020 Jun Epub: 2020 Apr 20 Authors Tao Y , Huang F , Zhang Z , Tao X , Wu Q , Qiu L , Wei H ,
Summary Publication Publication
Probiotic Potential and Safety Evaluation of Enterococcus faecalis OB14 and OB15, Isolated From Traditional Tunisian Testouri Cheese and Rigouta, Using Physiological and Genomic Analysis.
Frontiers in microbiology (Front Microbiol ) Vol: 10 Issue Pages: 881
Pub: 2019 Epub: 2019 Apr 24 Authors Baccouri O , Boukerb AM , Farhat LB , Zébré A , Zimmermann K , Domann E , Cambronel M , Barreau M , Maillot O , Rincé I , Muller C , Marzouki MN , Feuilloley M , Abidi F , Connil N ,
Summary Html Article Publication
Draft Genome Sequences of the Probiotic Enterococcus faecalis Symbioflor 1 Clones DSM16430 and DSM16434.
Genome announcements (Genome Announc ) Vol: 4 Issue 5 Pages:
Pub: 2016 Sep 29 Epub: 2016 Sep 29 Authors Fritzenwanker M , Chakraborty A , Hain T , Zimmermann K , Domann E ,
Summary Html Article Publication
Probiotic Enterococcus faecalis Symbioflor® down regulates virulence genes of EHEC in vitro and decrease pathogenicity in a Caenorhabditis elegans model.
Archives of microbiology (Arch Microbiol ) Vol: 199 Issue 2 Pages: 203-213
Pub: 2017 Mar Epub: 2016 Sep 21 Authors Neuhaus K , Lamparter MC , Zölch B , Landstorfer R , Simon S , Spanier B , Ehrmann MA , Vogel RF ,
Summary Publication Publication
In vitro comparison of the effects of probiotic, commensal and pathogenic strains on macrophage polarization.
Probiotics and antimicrobial proteins (Probiotics Antimicrob Proteins ) Vol: 6 Issue 1 Pages: 1-10
Pub: 2014 Mar Epub: Authors Christoffersen TE , Hult LT , Kuczkowska K , Moe KM , Skeie S , Lea T , Kleiveland CR ,
Summary Publication Publication
Impact of actin on adhesion and translocation of Enterococcus faecalis.
Archives of microbiology (Arch Microbiol ) Vol: 196 Issue 2 Pages: 109-17
Pub: 2014 Feb Epub: 2013 Dec 21 Authors Peng Z , Krey V , Wei H , Tan Q , Vogelmann R , Ehrmann MA , Vogel RF ,
Summary Publication Publication
Complete Genome Sequence of the Probiotic Enterococcus faecalis Symbioflor 1 Clone DSM 16431.
Genome announcements (Genome Announc ) Vol: 1 Issue 1 Pages:
Pub: 2013 Jan Epub: 2013 Feb 7 Authors Fritzenwanker M , Kuenne C , Billion A , Hain T , Zimmermann K , Goesmann A , Chakraborty T , Domann E ,
Summary Html Article Publication
In vitro comparison of commensal, probiotic and pathogenic strains of Enterococcus faecalis.
The British journal of nutrition (Br J Nutr ) Vol: 108 Issue 11 Pages: 2043-53
Pub: 2012 Dec 14 Epub: 2012 Feb 21 Authors Christoffersen TE , Jensen H , Kleiveland CR , Dørum G , Jacobsen M , Lea T ,
Summary Publication Publication
Comparative genomic analysis for the presence of potential enterococcal virulence factors in the probiotic Enterococcus faecalis strain Symbioflor 1.
International journal of medical microbiology : IJMM (Int J Med Microbiol ) Vol: 297 Issue 7-8 Pages: 533-9
Pub: 2007 Nov Epub: 2007 Apr 27 Authors Domann E , Hain T , Ghai R , Billion A , Kuenne C , Zimmermann K , Chakraborty T ,
Summary Publication Publication
Functional characterization of pro-biotic pharmaceuticals by quantitative analysis of gene expression.
Arzneimittel-Forschung (Arzneimittelforschung ) Vol: 53 Issue 5 Pages: 385-91
Pub: 2003 Epub: Authors Giese T , Zimmermann K , Meuer SC ,
Summary Publication Publication
[Reduction of acute recurrence in patients with chronic recurrent hypertrophic sinusitis by treatment with a bacterial immunostimulant (Enterococcus faecalis Bacteriae of human origin].
Arzneimittel-Forschung (Arzneimittelforschung ) Vol: 52 Issue 8 Pages: 622-7
Pub: 2002 Epub: Authors Habermann W , Zimmermann K , Skarabis H , Kunze R , Rusch V ,
Summary Publication Publication
[The effect of a bacterial immunostimulant (human Enterococcus faecalis bacteria) on the occurrence of relapse in patients with].
Arzneimittel-Forschung (Arzneimittelforschung ) Vol: 51 Issue 11 Pages: 931-7
Pub: 2001 Nov Epub: Authors Habermann W , Zimmermann K , Skarabis H , Kunze R , Rusch V ,
Summary Publication Publication

Now the gut!?!

Here we have few studies.

From the last study we see Veillonella, Corynebacterium 1, Romboutsia, Aerococcus and, Megasphaera, increasing, Lachnoclostridium decreasing. Fortunately, one of our tools on Microbiome Prescription allows us to enter these shifts and then compute with AI what would counter them.

The results are shown below (remember, they are scaled so the highest value is 1 — and the value is based on not the impact, but the number of studies reporting a desirable shift)

For those not familiar with fucoidan, Fucoidan is a long chain sulfated polysaccharide found in various species of brown algae. Commercially available fucoidan is commonly extracted from the seaweed species Fucus vesiculosus, Cladosiphon okamuranus, Laminaria japonica and Undaria pinnatifida [wikipedia]. It is available from many suppliers on Amazon, for example:

$22.00

Bottom Line

When I started working on this post, I was not expecting to fine anything significant. To one’s surprise, there was an ideal study published this year using human data with different degrees of severity. There are two really strong suggestions to be discussed with your MD before starting. – most of the rest come from Artificial Intelligence on the microbiome alone and not clinical experience; yet agree with existing studies.

Remember, these suggestions are coming solely from the reported bacteria shifts with no information about the medical condition using AI. The suggestions are supported by medical studies (where there are some)

Predicted Symptoms – Performance Review

At present we have many ways to estimate symptoms from a sample. In some cases, we look at what bacteria as a group produces, in other cases just the bacteria. This post is going to look at how well different method behaves. The different approaches are fishing expeditions to see if we can find more predictive analysis tools.

Method

Today, I added the ability to see predicted symptoms against entered symptoms, as shown below


We are going to pick the samples with the most symptoms, one per user and see how well each compares. Sample B had nothing from KEGG — KEGG is works from Species (and this sample lacked any). The nu,mbers below are matches in the top 20 predicted symtp,s/

Method (Top # selected)ABCDEFGHIJKLM
Consensus (30)11161169631067144
Bacteria (20)916912111212291441112
End Products (20)1181074765973611
KEGG Enzymes (20)8085833857134
KEGG Modules (20)130151058951071059
KEGG Products (20)701076710499367
Walking a collection of samples (each from a different person), All samples had at least 80 symptoms entered.

There are challenges with the symptoms entered – namely

What we do find is that every sample had at least one method identifying at least 50% of the person’s symptoms with the highest being 80%, followed by 75%,

Second we find that predicting from bacteria or KEGG Modules had the best performance. In no case was bacteria end products nor KEGG products nor KEGG Enzymes the best. Consensus only once matched the performance of the other two and was often very bad.

Bottom Line

The above shows a strong association of the bacteria (or it’s functions) to various symptoms. It does not prove causality, but causality is my working hypothesis (at least as a catalyst or contributor to symptoms).

This means that modifying bacteria may results in reduction or elimination of many symptoms.

Microbiome Review of a SIBO etc

Background

  • A nasal swab showed large amounts of staph aureas (but no MARCoNS) and k. oxytoca.
  • A stool test from Doctor’s Data that was done prior to SIBO treatment and treatment with a rotation of probiotics and prebiotics.
  • Partially hydrolyzed guar gum (PHGG) gave me a lot of bowel mucus after a few weeks, so I stopped it.
  • Tested with NirvanaBiome

Analysis

Because of the mention of SIBO and because I had just finished refactoring Symptom to Bacteria association, I decided to start there. I do not have great expectations because published clinical studies were unable to find any significant patterns.

See Symptoms Associations post for how we are doing this

From https://www.microbiomeprescription.com/Explorer/ToSymptomsBacteriaSummary

We now see hoe well this person matches..

As expected, no significant associations

Looking at Other Explorers

  • KEGG Enzymes – nothing
  • KEGG Module:33. Very Strong: 29, Strong: 4 – this is interesting… especially since all 33 are Middle Peaks. This should be explore in time. For anyone interested in exploring,
    • beta-Oxidation, acyl-CoA synthesis
    • Biotin biosynthesis, pimeloyl-ACP/CoA => biotin
    • CAM (Crassulacean acid metabolism), light
    • Citrate cycle (TCA cycle, Krebs cycle)
    • Citrate cycle, second carbon oxidation, 2-oxoglutarate => oxaloacetate
    • CMP-KDO biosynthesis
    • Cobalamin biosynthesis, cobyrinate a,c-diamide => cobalamin
    • D-Galacturonate degradation (bacteria), D-galacturonate => pyruvate + D-glyceraldehyde 3P
    • Gluconeogenesis, oxaloacetate => fructose-6P
    • Glycogen degradation, glycogen => glucose-6P
    • Glycolysis (Embden-Meyerhof pathway), glucose => pyruvate
    • Glycolysis, core module involving three-carbon compounds
    • Histidine biosynthesis, PRPP => histidine
    • Histidine degradation, histidine => N-formiminoglutamate => glutamate
    • Inosine monophosphate biosynthesis, PRPP + glutamine => IMP
    • Isoleucine biosynthesis, pyruvate => 2-oxobutanoate
    • Leucine biosynthesis, 2-oxoisovalerate => 2-oxoisocaproate
    • Lysine biosynthesis, DAP dehydrogenase pathway, aspartate => lysine
    • NAD biosynthesis, aspartate => quinolinate => NAD
    • Pantothenate biosynthesis, valine/L-aspartate => pantothenate
    • Pentose phosphate pathway (Pentose phosphate cycle)
    • Pentose phosphate pathway, non-oxidative phase, fructose 6P => ribose 5P
    • Pentose phosphate pathway, oxidative phase, glucose 6P => ribulose 5P
    • Phosphate acetyltransferase-acetate kinase pathway, acetyl-CoA => acetate
    • Phosphatidylethanolamine (PE) biosynthesis, PA => PS => PE
    • Pimeloyl-ACP biosynthesis, BioC-BioH pathway, malonyl-ACP => pimeloyl-ACP
    • Pyrimidine ribonucleotide biosynthesis, UMP => UDP/UTP,CDP/CTP
    • Riboflavin biosynthesis, plants and bacteria, GTP => riboflavin/FMN/FAD
    • Serine biosynthesis, glycerate-3P => serine
    • Tetrahydrofolate biosynthesis, GTP => THF
    • Tryptophan biosynthesis, chorismate => tryptophan
    • Uridine monophosphate biosynthesis, glutamine (+ PRPP) => UMP
    • Valine/isoleucine biosynthesis, pyruvate => valine / 2-oxobutanoate => isoleucine
  • KEGG Product:50. Very Strong: 36, Strong: 5, Weak: 6, Very Weak: 2

Adjusting for Middle Peak

Finding middle peaks presents some challenges for suggestions. The graphics below may help explain with a middle peak is. The why is not simple, but likely a complex interaction of many things (like other bacteria)

But looking at the raw numbers will have most people seeing “nothing”

KMFunction(value) often reveal relationships that are hard to see in the original numbers

As a result of doing this post, I realized that it was possible to generate suggestions for these middle peaks. This is explained in the video below


The resulting suggestions are below. These are very blinkered suggestions that ignore extreme values.

Consensus Suggestions

See Multiple Conditions Microbiome Analysis post for more background.

I then did two common suggetions

  • Kaltoft-Moltrup Suggestions
  • Jason Hawrelak

This results in 3 items in the Consensus

KEGG Suggestions

The clear winner is Sun Wave Pharma/Bio Sun Instant probiotics – consisting of Clostridium butyricum and Bacillus mesentericus.

For supplements, we have

  • beta-alanine
  • iron
  • L-Phenylalanine
  • L-Tyrosine
  • Vitamin B-12

Medical Conditions(PubMed) had only one item that was border line item, Graves’ disease (an immune system disorder that results in the overproduction of thyroid hormones (hyperthyroidism). A simple blood test should clarify this risk.

Thanks very much – Based on a quick skim, it’s interesting that Graves comes up bc my dad has it and my TSH has been declining the last couple of times it was measured, but they didn’t check my T4, so I need to have another test soon that will look at actual hormone levels.

From person after reviewing the draft,

Putting it all together – Consensus

On my first drafts of this post, I did the manually and then realized automating it was a much better solution for me and for others.

REMINDER: The numbers are NOT the degree of impact. It is the confidence that it will help. That is, the number of bacteria shifted in the right direction, the number of studies finding the shifts and the amount of shifts we want to see.

:

There is now a better way!

I have enabled a Consensus Report on all suggestions done on a specific sample in the last 24 hours. After 24 hours, I remove the data, First, I will do the two most common suggestions:

After the 2nd one, a new choice appears on the home page sample menu as shown below

New item appears when there are two or more. You can remove them and start over as desired

I then added Comorbid: Small intestinal bacterial overgrowth (SIBO) to the mixture,

I ahd tried a fourth one, I went to Advance Suggestions and selected PubMed for SIBO

Unfortunately, there were NO MATCHES to this data source (Not a surprise)

Now, let us look at the Consensus View, it’s divided into 6 sections as shown below

NOTE: by default it is sorted by name. Double click the value column and then double click the Count column will sort it as shown below (i.e. best or worst at top of list).

Recap

This was an interesting analysis in that PubMed literature failed to find any matches, while citizen science found patterns. The patterns were often middle peaks which would not have been seen with cookbook statistical methods focused on normal distributions (bell curves)

Suggestions 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.

One of the confusing aspects of Microbiome Prescription is that suggestions are derived through different paths with suggestions from different paths being in partial contradiction. Each path has limited and often fuzzy data. The safest path is to merge the lists with what is in common, the next step forward (if the first step does not give you too much to do), is items that is cited in one list and not contradicted in another list.

Often what suggestions result in

Multiple Conditions Microbiome Analysis

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.

.

Adjusting when our world view is a simplistic “too high” or “too low”

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

Gerd does not have many samples
Crohn’s has two cases — since more samples means more accurate detection, we will use the highest one only
With Mast Cell, we see that the relationships goes down as sample size goes up — we are likely getting better resolution

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.

Consensus Report persists for only 24 hours or until you manually clear them

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.

Increasing the number above will reduce the size of the list. Decreasing will increase the size

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

This person is already does broccoli etc regularly —
Broccoli shows up again — a duplicate record issue that I will be fixing soon
A very short list
Note that sea weed is there BUT a different one — oh the fine details!

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.

Walk thru of earlier version – minor changes

Symptoms Associations

Over the last few years, I have been trying to tease relationship out of data. I have tried a wide variety of methods and finally found one that been producing good results.

The method is conceptually straight forward:

  • Take the actual reading and apply a monotonic increasing function to it. Thus if Valuea < valueb then func(Valuea) < func(valueb)
  • With the resulting data, transform it to be a rectangular distribution for all samples
  • Hypothesis test the values from people who recorded symptoms using P=0.01 as a threshold

Once the candidate association are done then we can also test if a sample’s item satisfies the hypothesis.

This approach has some nice characteristics, because it will detect patterns that:

  • are not linear on the values
  • does not assume a normal distribution
  • does not not assume items are caused by end associations (i.e. too high or too low)
    • In some cases, we see a shift into a middle range that is statistically significant

Adjusting “Middle Peak” patterns

Both of the above above are typical beliefs that people will attempt to apply to the data.

Probability distribution function; (a) uniform distribution of the... | Download Scientific Diagram
Comparing uniform distribution to normal distribution

Seeing the Bacteria Interactions in Your Sample

I have refactored Bacteria Interactions Why? on the site to use the data discovered via this post. The information is more accurate and more comprehensive than the prior version

I have done a quick demo, shown below, and I will add a few examples after.

Key points

  • The color of the oval indicate level of hierarchy
    • Same color to the middle one indicates that it’s independent
    • Often other colors are children or parents
  • Line thickness indicate amount of influence
  • Size of the ovals indicate percentile ranking.
    • A small oval indicates less than bacteria than normally seen
    • A large oval indicates more bacteria than normally seen

Examples

In our first example below we see many other species encourages this species If we look at it’s hierarchy on NCBI, we see a lot of bacteria that are not related by descent.

 Firmicutes; Negativicutes; Veillonellales; Veillonellaceae; Veillonella

Chart of Species Veillonella parvula

For the next example we see many siblings influencing it.

Bottom Line

Beyond the fun to see aspect, if there is an item of concern you may wish to see what bacteria influences it and include those in a hand-picked sample.

A Fecal Matter Transplant goes wrong (several cases)

A reader sent me an email shown below:

…did a NirvanaBiome test before and 4 weeks after finishing the FMT. As far we can see her gut condition only got worse.”

On The Gut Club: Stool Test Discussion Group Facebook group, some other comments were shared

My first goal is to try to understand what went wrong and how. Then looking at “where do we go from here”. Ending with a reading list on FMT.

Additional Personal Experience with FMT shared

Personal Email (with permission)
From Facebook

Analysis of Changes

I started with comparing Bacteria/Taxonomy Out of Range Over Time which showed 63 items. The most interesting are below. Percentile means where in a collection of 2000+ samples that the reading is. For example 97%ile means that 60 samples had more than this and 1940 samples has less – most people would deem that to be an excessive overgrowth. Prior means before the FMT, After means weeks after the FMT.

BacteroaPrior PercentileAfter Percentile
(class) Clostridia97.899.3
(class) Gammaproteobacteria95.4none
(family) Lachnospiraceae95.8none
(family) Ruminococcaceaenone100
(genus) Ruminococcus99,8100
(order) Enterobacterales97none
(species) Ruminococcus bicirculansnone100%
Nirvana data structure has some issues converting to string NCBI hierarchy/

My initial thought on seeing these results was – are we using the same lab? We are.

The mechanism of getting names is finding patterns in the raw data. Lachnospiraceae and Ruminococcaceae are both children of Clostridiales, so we may have

This dramatic shift may be (1) a defect in the classification algorithm or (2) a bug in my specific import routine for CosmosId [which I am looking at) or (3) sibling families taking over [siblings tend to like the same environments] , I am inclined to the first cause(1) since the data is pushed through the same code(2) but (3) is almost as probable as (1). See my 2019 The taxonomy nightmare before Christmas… post for background.

Pie Charts

The two charts below show the growth of Clostridiales and the reduction of every other order. I have often described Fecal Matter Transplants as equivalent to an Organ Transplant (or Blood Transfusions) with the same issues of rejection being significant. We do not know yet now to “type” or test the new item for compatibility.

I speculate that there may have been warfare with several orders weakened, Clostridiales was already near an extreme value — my gut feeling is that bacteria with strong overgrowth are dominated by strains that have the following characteristics

  • They are more robust (i.e. more resistant to bacteriocins (natural antibiotic) produced by other bacteria
  • They produce significantly more strong bacteriocins
PRIOR: Clostridiales is 54% of Bacteria
AFTER: Clostridiales is 94% of Bacteria

The diagram below helps to explain bacteriocins. It show different strains of Lactobacillus Reuteri. Reuterin is the bacteriocins(natural antibiotics) produced. Some produce a high amount, some a low amount and others none.

Reuteri.PNG
From prediction to function using evolutionary genomics: human-specific ecotypes of Lactobacillusreuteri have diverse probiotic functions[2014].

Drilling down into Clostridiales we see major shifts. CosmosId report by strains, so these numbers are reliable

PRIOR
AFTER

What we see ins that one species EXPLODED, Ruminococcus bicirculans. Prior to the FMT, it’s count was 47,810 or 0.5%, after it jumped to 444,900 (44.5%!!!) – a 10x increase. Whether it was a new strain that was introduced by the FMT that the existing microbiome could not handled as well as the donor (I deem more likely) or an existing strain that filled the vacuum resulting from the FMT bacteria and native bacteria wiping each other out (less likely) is speculation.

Suggestions

We have an idea of what may happen, apart from being a visual warning in the above diagrams to people considering FMT. The key question for this person is how to undo it. For this troublesome strain, we know some things that will increase or decrease it. I would focus on this strain alone and do a deep “spring cleaning” of supplements.

Looking at the Kaltoft-Moltrup Range based suggestions, we see a lot of bacteria selected as abnormal

On AFTER sanoke

And we see most of the above items are on the suggestions for this combination:

Bottom Line

I would suggest (after reviewing with your medical professional) the above changes and then do another Nirvana sample after 6-12 weeks to see what has changed. I would expect this Ruminococcus bicirculans. to be significantly reduced, the question is what will replace the 44% of the microbiome it currently occupies.

I was curious if we could infer something about the donor by looking at these changes. I looked at life style that could impact Ruminococcus bicirculans. etc and found that

  • INCREASE:  reduced calories low-carbohydrate high-fat diet [2021]
  • INCREASE:  exercise-stress [2016]

A common sense (A priori) definition of what would be a healthy donor will often be someone that fits these two features, i.e. an athlete. ” Recent trends show more athletes trying a low-carbohydrate, high-fat (LCHF) diet for endurance performance. ” [Low vs. High Carbohydrate Diets for Endurance Performance] IMHO, this a priori common sense healthy donor is wrong — athletes have abnormal microbiomes. A priori is defined as “denoting reasoning or knowledge which proceeds from theoretical deduction rather than from observation or experience.“

The compatibility issue is not only at the microbiome but diet and exercise style. If the recipient does not consume the same diet as the donor, the transfer may go in odd directions. Similarly, we know exercise impacts the microbiome. A “desk jockey” getting the microbiome of someone that cycles for 4 hours a day will likely fade quickly.

Prior Posts on FMT

My interest in trying to understand different FMT responses go back at least 5 years. The following posts are likely worth reviewing. My feeling continues to be that the downside risk for the upside benefit is still too low to be a desirable course of action. Typically, it is an improvement that does not persist for extended periods. I believe a committed microbiome manipulation with regular retests and adjustment has considerably less downside and equivalent upside benefit that will likely persist longer — especially, if at least two samples and adjustments are done a year once sufficient benefit has been obtained.

Different ways of Doing FMT

I am tossing some technical information for those that are interested in how FMT can be done. There are many ways, including “the turkey baster” aka “Turd Burglars“. IMHO, it should be done under medical supervision with adequate testing (including comparing the donor’s microbiome, and life style to the recipient)

By Enema

This was used by the above

FMT with Enema – 6 doses, 1 every other day (11 days)

No antibiotics are used!

Pre- and probiotics (phgg, bifido)


Prior

  • Do not stop medication
  • follow a liquid diet the night before the first administration (and not the other 5!!!)
  • cleansing water enema in the morning before FMT
  • place the dose from the freezer in the refrigerator the night before. allow the dose to reach room temperature in the morning.
  • swallow 1 imodium after awakening (only) for the first dose
  • limit yourself to a light (liquid) meal

During

  • do not eat or drink anything during and immediately after the enema.
  • work in stages with a limited dose (but within 10 minutes on the left side)
  • use gravity to keep everything inside, e.g. a pillow under the hip
  • take a few deep breaths regularly
  • alternate position (left side, back, stomach, right side), every ten minutes (left is best for keeping inside)
  • regularly massage the dose upwards gently
  • try to keep the dose for ideally more than 6 hours
  • rest, relaxation and sleep are the best activities during the therapy days

After

  • do not immediately make major changes to your daily diet
  • not drinking and being (somewhat) thirsty helps to withdraw fluid from the injected dose
  • fibers help the new bacteria to settle. gradually introduce fiber into your diet. fiber rather through food than supplements.

Centre for Digestive Diseases – Doctor Thomas Borody

” Dr Borody has overseen over 12,000 FMTs, creating a wealth of proprietary clinical data and insights.” Antibiotics are frequently used prior to FMT. He is likely the leading Australian MD that does FMT.

  • Faecal Microbiota Transplantation
    • ” published response rates in some studies using FMT to treat UC greater than 50%1.”
    • “In our experience FMT may help symptoms of Irritable Bowel Syndrome (IBS) however, this not guaranteed. “

Selected publications

There are more publications on his website.

Bacteria influencing other Bacteria

Recently I revisited finding association between bacteria. We know bacteria both produce and consume metabolites and chemicals, as well as bacteriocins that will inhibit other bacteria. “Bacteriocins are potential alternatives to traditional antibiotics. These peptides, which are produced by many bacteria, can have a high potency and a low toxicity” {Nature 2012]. Finding the relationships has been a challenge because of the nature of the distribution (not a bell curve — see this post on the solution that I use for identifying abnormal values) Post #1 Post #2.

This is a technical note (WARNING: Geek Speak) on the 262,603 relationship with correlation coefficient R2 of 0.10 or higher that is available on the site.

Example of Classic Association

For our example, we will compare two families: Brucellaceae and Caulobacteraceae. Their ancestry is shown below

Because they have some shared ancestry, you would usually expect them to be friendly and suppurative of each other. The standard analysis is shown below, charting the counts from samples that have both bacteria.

Classic Approach

After an intro course to statistics, most people would do a regression. It is unlikely they would look at the chart because there are 2,669,956 charts that would be produced with the dataset that I am working with.

The regression and the chart is shown below, logical conclusion – no relationship.

Each axis is the count of a specific bacteria (Brucellaceae and Caulobacteraceae) from the same sample.

Alternative Approach

The alternative is to use what is called monotonic increasing functions on the counts. We scale the function so that it’s range is 0 to 100. This preserves the nature of the data and discard the naïve assumption of linearity. The result is shown below. With this approach, we get the following chart. same data!!!

Each axis is the transformed count of a specific bacteria (Brucellaceae and Caulobacteraceae) from the same sample.

We could for each pair of bacteria derive the absolute optimal monotonic functions. This approach I find problematic because your appear to be fiddling with the data too much. I have put the additional constraint that you are allowed only one monotonic function per bacteria. I believe this will inhibit over-fitting the data to the model.

How many relationships over 0.1?

We have 1621 bacteria with at least one, and the top ones are shown below

taxonomy ranktaxonomy nameCount
familyHalanaerobiaceae546
classFibrobacteria526
classDehalococcoidia506
familyFibrobacteraceae505
orderFibrobacterales501
genusFibrobacter483
familyNitrosomonadaceae474
genusDehalogenimonas474
orderAcidobacteriales467
familyMicromonosporaceae461
genusNitrosomonas460
genusAcinetobacter459
familyColwelliaceae455
familyAcidobacteriaceae453

What benefit does this give?

The impact of one bacteria on the other may be computed as slope * r2 . So R2 of .5 and a slope of .4 = .5 * .4 = .20 or 20%, thus for every 10 steps of one, the other will increase by 2.

We can use this when some bacteria X is high or low and we have no information on modifying it. We can look at the related bacteria with highest impact and its modifiers. We are trying to cascade by changing the associated bacteria to change our target bacteria! We are attempting to model the modifiers secondary changes into our suggestions.

Where is this on the site?

On the bacteria details pages. if there are associations, there will appear a link to it

Clicking this will take you to the impact page. In the example below you see that Lactobacillus accounts for 63% of it’s parent class. Lactobacillaceae(family) which includes  Lactobacillus , Pediococcus  , and    Sharpea. So it is the greatest contributor the three.

Orphan Detail Pages

I call these orphan because there is not literature on them or little studies. For example Pectinatus where there was just one know citation, ginko. We now have 10 more marked with the association icon as shown below.

Available to include for Suggestions

There is a new checkbox on the custom suggestion page. If you wish these to be factored into suggestions just check the box.

Reviewing the impact of Ivermectin and Nystatine

A reader asked for a review. The reader had a prior sample taken 6 weeks before and specific treatments between

I took Ivermectin  for 4 days during this month – one per day 12mg. I am treating yeast with nystatin 5,000 units 2 times per day for one month. I hope the die-off may have made room for bacteria to grow. I still feel crappy. I stopped lactobacillus ( I tried Lactobaciullis grains from Keith during that time. too much bloating), and started Akkermansia muciniphila and very recently some Bifidobacterium. I had cut back on dairy earlier (using soy milk and oat milk instead). I also took Sporonax (Itraconazole) antifungal about 5 times and Valtrex 500 mg 8 times.

My starting point (before looking at the samples) is to look at what we know about the impact of these items.

So the expectation of making room for bacteria appears very reasonable. The unfortunate aspect is that among the causalities are: Lactococcus, Lactobacillus, Bifidobacterium and Akkermansia muciniphila. So the question arises, will the good or the bad grow back faster?

What changed?

I first checked the common bacteria that most people are usually concerned with (cited above) and then will look at what increased. One item is of definite concern 💥, (class) Fusobacteria

BacteriaBeforeLatest
(genus) Lactobacillus1150⇲140
(genus) Lactococcus380⇲180
(genus) Bifidobacterium640⇲340
(species) Akkermansia muciniphila50⇲30
(species) Hathewaya histolytica220⬆️869
(class) Negativicutes10090⬆️14820
(class) Gammaproteobacteria790⬆️1500
(class) Bacteroidia267,859⬆️322,050
(class) Chitinophagia20⬆️70
(class) Fusobacteria55280⬆️90210 💥
(class) Coriobacteriia3160⇲670
(class) Sphingobacteriia1010⇲270

Checking back with what is decreased by the drugs, at the family level

It was interesting to note the many of the bacteria that were abnormally high (95%ile) stayed the same or increased. The fact that they were high implies more aggressive strains (and possibly more bacteriocin and antibiotic resistant).

Looking at species that are outside of the KM range, we have the following being excessively high

Note that other Blautia species were abnormally low. These are already accounted for in the suggestions.

Looking further back

“I believe vancomycin started the problem to flare when I took in a couple of years ago. I also had my first covid shot of June 26, 2021 with immediate bad reactions and probably made worse since.

The very high Fusobacteria identified above is one bacteria that would be decreased by this antibiotic (so this is unlikely that this the cause of this being high).

The following are items that have been reported to increase this bacteria:

Suggestions

I expect the suggestions to be very similar because the items high before stayed high (i.e. no change)

Suggestions were done with the Kaltoft-Moltrup ranges.
Latest Suggestions
Suggestions from earlier sample

There was one additional probiotic suggested from the latest sample and it had a far stronger impact then the older ones.

Last SampleL
Prior Sample

The KEGG suggested probiotics are the same ones, just a higher value because of an increased deficiency in enzymes being produced by the microbiome.

Latest Sample
Prior Sample

Bottom Line

The data base correctly predicted the likely decrease of the common bacteria assumed important for health. Those that did no reduce were either not effected, more resistant strains (inferred from being high numbers prior)

What I found interesting was the absence of most probiotics (lactobacillus and bifidobacterium) in the safe suggestions, except for bifidobacterium breve. Securil (Propionibacterium freudenreichii) probiotic came in very strong.

  • Giloteaux65 found that supplementation with the bifidogenic substance Propionibacterium freudenreichii improved butyrate levels, which induces an anti-inflammatory cascade [2017]
  • One of particular interest is found in the product Securil and is called – Propionibacterium freudenreichii, which produce propionic acid, a natural biological acid that benefits the bifidus flora. [2010]
  • More studies here

Note that this is specific for this person. Suggestions will be different for other peoples taking the same items.

Reader’s Desire

“I was also trying to get my methylation working better and taking b vitamins again. Could it be the b vitamins?”

The B-vitamins are suspect for the high levels of Fusobacteria. We should also note that are the to avoid list we see many of the B vitamins which suggests that much of the dysbiosis may be vitamin B related.

Concerning methylation, I usually see what enzymes are involved by checking with the Kyoto Encyclopedia of Genes and Genomes, And then check the KEGG Enzyme Outliers report. This report had a very high 470 items listed!!

A few of outliers showing extreme values

Drilling down into a few, we see some of the causes

This value is beyond the K-M range and thus suggestions to correct this is automatically captured in the above suggestions.

Example of Microbiome manipulation for hypertension

After a recent hospital visit for cellulitis (with many different antibiotics, both orally and by IV), my blood pressure was significantly elevated that the substitute MD (my usual was on vacation), that I was put on Lisinopril. Within a week I developed a dry cough that has for 35 years has been a “tell” for a relapse into Myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS). Checking the literature, I found that about 30% of people develop this cough. To me it is an important tell, if it shows up — I need to do quick re-examination of what is going on. For a prescription drug to do so, really made me uncomfortable.

I then check Lisinopril against the bacteria shifts reported for ME/CFS, and it made them worst. In short, staying on it may well increase the risk of relapsing into ME/CFS. That is not acceptable.

I did a little more research and found a variety of different opinions on COVID and the use of ACE2, one example is Long-term ACE Inhibitor/ARB Use Is Associated With Severe Renal Dysfunction and Acute Kidney Injury in Patients With Severe COVID-19: Results From a Referral Center Cohort in the Northeast of France, 2020

In 2019, I had done a posting on hypertension citing Nutrients and Nutraceuticals for the Management of High Normal Blood Pressure: An Evidence-Based Consensus Document. [2019].

I stopped the lisinopril and proceeded to take the nutrients etc cited above, at or above the specified dosages. I know that it will take a little time for the microbiome to respond, but it did.

The spikes were from not waiting long enough after exercising. I now rest at least 20 minutes

Bifidobacterium Correlation

Reviewing the literature, there is the appearance of blood pressure being strongly associated with the amount of bifidobacterium as we age. Children are very high in Bifidobacterium and low in BP. As the typical amount of bifidobacterium decreases with age, blood pressure increases.

I found this recent study,

As a result, I add 2 tablespoons of bran to the typical 4 table spoons of oats porridge that was doing. I also added a package of Holigos (Human Milk Oligosaccharides)  which I know is a super feeder of bifidobacterium.

This corresponded to the severe drop shown above.

Possible Probiotic Impact

I was taking the following based on modelling of the bacteria shifts seen in hypertension:

Samples are coming…

Just got notification from the lab that last weeks was received.

One addendum, when I was in hospital for cellulitis, my potassium was very low and I required a (painful) IV of potassium. I examined the amount of potassium that my usual diet provided… It was very low, so I started to supplement with potassium citrate also.

Bottom Line

I was able to normalize (for an almost 70 year old) blood pressure by using existing research and having patience. I believe the key items was encouraging bifidobacterium growth (sorry, bifidobacterium probiotics do not persist usually and have little impact), correcting mineral content (potassium, magnesium, calcium).

One more addendum, I usually did 10,000 steps a day with weekend hikes often being as high as 20,000 steps.