Special Studies: Depression

This is a common symptom for our times, especially with climate change and politics in some places. This is reported often in samples, and thus being examined if it reaches our threshold for inclusion as defined in A new specialized selection of suggestions links. It does.

Depression is likely a part of a feedback loop — depression chemicals modify the microbiome resulting in more depression triggering chemicals being released.

Study Populations:

We have 2 symptom annotations that could be included

  • General: Depression
  • Official Diagnosis: Depression

Taking the first two together we get 85 samples with an max z-score of 5.7

SymptomReferenceStudy
High Anxiety and Anxiety/tension114985
  • Bacteria Detected with z-score > 2.6: found 156 items, highest value was 7.8
  • Enzymes Detected with z-score > 2.6: found 508 items, highest value was 8.3
  • Compound Detected with z-score > 2.6: found No items

This hints that the enzymes being produced are likely more significant than the bacteria and the number of enzymes found significant is slightly overwhelming! (I dare not say depressing)

Interesting Significant Bacteria

All bacteria found significant (except 2) had too low levels. These two are almost expected as the false detection rate, so we can likely exclude them. This is a common pattern found with these studies, it is not “bad bacteria bogie man bacteria” but an absence of “upstanding citizens bacteria”.

BacteriaReference MeanStudyZ-Score
Shuttleworthia (genus)273427.8
Actinobacillus (genus)339806.6
Actinobacillus porcinus (species)181486
Haemophilus (genus)15984255.9
Haemophilus parainfluenzae (species)15984245.9
Pasteurellaceae (family)19735345.8
Pasteurellales (order)19735345.8
Pasteurellaceae incertae sedis (norank)143435.8
[Pasteurella] aerogenes-[Pasteurella] mairii-[Actinobacillus] rossii complex (species group)143435.8
Legionellaceae (family)80405.7
Legionella (genus)80405.7
[Actinobacillus] rossii (species)143465.6
Legionellales (order)81435.2
Thiobacillus thiophilus (species)81245.2
Veillonella dispar (species)823595.1
Thiobacillaceae (family)81255
Thiobacillus (genus)80255

Similar findings have been reported in some studies, for example: low Haemophilus, Pasteurellaceae, uncultured_Veillonellaceae_bacterium reported in Gut Microbiome Composition Associated With Major Depressive Disorder and Sleep Quality [2021].

Interesting Enzymes

All 508 enzymes found significant, had too low levels. Excuse the long list, but my practice has been to list everything with a z-score over 5.0; there are a lot with depression.

EnzymeReference MeanStudy MeanZ-Score
[cysteine desulfurase]-S-sulfanyl-L-cysteine:[molybdopterin-synthase sulfur-carrier protein]-Gly-Gly sulfurtransferase (2.8.1.11)523615178.3
tRNA-uridine65 uracil mutase (5.4.99.26)40599717.6
deoxyribocyclobutadipyrimidine pyrimidine-lyase (4.1.99.3)415611796.8
acyl-CoA:sn-glycerol-3-phosphate 1-O-acyltransferase (2.3.1.15)32709766.7
7,8-dihydroneopterin 3′-triphosphate diphosphohydrolase (3.6.1.67)360910846.6
(1->4)-alpha-D-galacturonan lyase (4.2.2.2)23527066.4
N4-acetylcytidine amidohydrolase (3.5.1.135)29708426.4
diacylglycerol-3-phosphate phosphohydrolase (3.1.3.4)32879206.3
1,2-diacyl-sn-glycerol 3-phosphate phosphohydrolase (3.1.3.81)32879206.3
acetyl-CoA:[elongator tRNAMet]-cytidine34 N4-acetyltransferase (ATP-hydrolysing) (2.3.1.193)33099386.3
S-adenosyl-L-methionine:tRNA (cytidine32/uridine32-2′-O)-methyltransferase (2.1.1.200)33129636.3
S-adenosyl-L-methionine:23S rRNA (uracil747-C5)-methyltransferase (2.1.1.189)32609316.3
protein dithiol:quinone oxidoreductase (disulfide-forming) (1.8.5.9)355511176.3
ATP phosphohydrolase (ABC-type, thiamine-importing) (7.6.2.15)35069736.2
ATP:(Kdo)-lipid IVA 3-deoxy-alpha-D-manno-oct-2-ulopyranose 4-phosphotransferase (2.7.1.166)15143806.2
[50S ribosomal protein L16]-L-Arg81,2-oxoglutarate:oxygen oxidoreductase (3R-hydroxylating) (1.14.11.47)33289786.2
D-glucarate hydro-lyase (5-dehydro-4-deoxy-D-glucarate-forming) (4.2.1.40)29589976.1
ATP:N-acetyl-D-glucosamine 6-phosphotransferase (2.7.1.59)32579476.1
fatty acyl-[acyl-carrier protein]:alpha-Kdo-(2->4)-alpha-Kdo-(2->6)-(acyl)-[lipid IVA] O-acyltransferase (2.3.1.243)33189606.1
[RNA] 5′-hydroxy-ribonucleotide-3′-[RNA fragment]-lyase (cyclicizing; [RNA fragment]-3′- nucleoside-2′,3′-cyclophosphate-forming and hydrolysing) (4.6.1.19)15153906.1
ATP:N-acyl-D-mannosamine 6-phosphotransferase (2.7.1.60)32049326.1
n/a (3.4.11.23)33059666.1
coproporphyrinogen:oxygen oxidoreductase (decarboxylating) (1.3.3.3)31538246.1
D-amino acid:quinone oxidoreductase (deaminating) (1.4.5.1)24846816.1
gamma-L-glutamyl-L-cysteinyl-glycine:spermidine amidase (3.5.1.78)31859406
gamma-L-glutamyl-L-cysteinyl-glycine:spermidine ligase (ADP-forming) [spermidine is numbered so that atom N-1 is in the amino group of the aminopropyl part of the molecule] (6.3.1.8)31859406
CMP-N-acetyl-beta-neuraminate:beta-D-galactosyl-(1->4)-N-acetyl-beta-D-glucosaminyl-R (2->3)-N-acetyl-alpha-neuraminyltransferase (configuration-inverting) (2.4.99.6)16034066
galactarate hydro-lyase (5-dehydro-4-deoxy-D-glucarate-forming) (4.2.1.42)30148865.9
methanesulfonate,FMNH2:oxygen oxidoreductase (1.14.14.34)22375115.9
alkanesulfonate,FMNH2:oxygen oxidoreductase (1.14.14.5)22375115.9
donor:hydrogen-peroxide oxidoreductase (1.11.1.21)30197975.8
D-glucose:ubiquinone oxidoreductase (1.1.5.2)21964975.8
glutathione:hydroperoxide oxidoreductase (1.11.1.27)17534135.8
3-hydroxybutanoyl-CoA 3-epimerase (5.1.2.3)29408265.7
acetyl-CoA:glyoxylate C-acetyltransferase [(S)-malate-forming] (2.3.3.9)31099935.7
L-methionine:2-oxo-acid aminotransferase (2.6.1.88)27497455.7
S-adenosyl-L-methionine:23S rRNA (guanine2069-N7)-methyltransferase (2.1.1.264)568526085.7
ferredoxin:NAD+ oxidoreductase (1.18.1.3)19904815.6
choline:acceptor 1-oxidoreductase (1.1.99.1)27396555.6
n/a (3.1.25.1)15434425.5
isocitrate glyoxylate-lyase (succinate-forming) (4.1.3.1)29428595.5
acetyl-CoA:dTDP-4-amino-4,6-dideoxy-alpha-D-galactose N-acetyltransferase (2.3.1.210)27627455.5
(S)-3-hydroxyacyl-CoA:NAD+ oxidoreductase (1.1.1.35)336110735.5
N-acyl-D-amino acid amidohydrolase (3.5.1.81)17874125.4
2-(glutathione-S-yl)-hydroquinone:glutathione oxidoreductase (1.8.5.7)31249165.4
S-(hydroxymethyl)glutathione:NAD+ oxidoreductase (1.1.1.284)28508925.3
succinyl-CoA:3-oxo-acid CoA-transferase (2.8.3.5)13063195.3
n/a (3.4.23.51)501623115.3
L-2,4-diaminobutanoate carboxy-lyase (propane-1,3-diamine-forming) (4.1.1.86)21203825.3
formate:[oxidized hydrogenase] oxidoreductase (1.17.98.4)599029165.3
S-methyl-5-thio-D-ribulose 1-phosphate 1,3-isomerase (5.3.3.23)1135955.1
RX:glutathione R-transferase (2.5.1.18)296110215.1
n/a (3.4.24.74)427911325.1
gamma-L-glutamyl-L-cysteine:glycine ligase (ADP-forming) (6.3.2.3)28089015.1
D-glucose:NAD(P)+ 1-oxidoreductase (1.1.1.47)12433095.1
propanoyl-CoA:phosphate propanoyltransferase (2.3.1.222)17254175
(S)-2-hydroxyglutarate:quinone oxidoreductase (1.1.5.13)21906015
4-hydroxyphenylpyruvate:oxygen oxidoreductase (hydroxylating, decarboxylating) (1.13.11.27)14372555
thiosulfate:ferricytochrome-c oxidoreductase (1.8.2.2)13613365

This high count is actually reflected in recent literature, although it was with animal models and not humans (in our case).

 Moreover, the fecal metabolomic analysis identified 468 differential expressed metabolites. Among all the differential metabolites, 11 specific pathways significantly altered, which were mainly belonged to lipid and amino acid metabolism.

Comparative analysis of gut microbiota and fecal metabolome features among multiple depressive animal models [2022]

Given the huge numbers of enzymes, I decided to see what came up from KEGG calculations for appropriate probiotics. If you have depression, the site will compute those specific for yourself. As seems to happen often, E.Coli produces a lot of different enzymes and thus at the top of the list.

Tax_NameContribution
Escherichia coli200.9
Bacillus thuringiensis70.2
Bacillus licheniformis67
Bacillus subtilis64.5
Bacillus subtilis subsp. natto63
Bacillus pumilus58
Bacillus velezensis57.8
Bacillus amyloliquefaciens56.8
Brevibacillus laterosporus45
Lactobacillus plantarum subsp. plantarum31.5
Lactiplantibacillus plantarum31.4
Clostridium beijerinckii28.3
Lacticaseibacillus rhamnosus23
Lacticaseibacillus casei23

Bottom Line

The high number of statistically significant enzymes was a surprise — but checking the literature, it appears to be in agreement with other studies. The special study suggestions engines are available now. Either via the a priori suggestion or the suggestions specific for your microbiome.

Given these facts, fixing depression may be a challenge until someone develops a Shuttleworthia and a Actinobacillus probiotic.

https://microbiomeprescription.com/Library/CitizenScience

Refactored KEGG based Probiotics

One thing about dealing with the microbiome is the need to constantly check and test your assumptions and revised from the results of tests. The series of A new specialized selection of suggestions links shattered several reasonable assumptions that appears to be false for most people, namely:

  • A surplus of specific bacteria is not the cause (no statistical significance was found)
  • A surplus of specific enzymes is not the cause (no statistical significance was found)
  • Abundance or deficiency of certain compounds rarely has any statistical significance.

What keeps showing up as statistically significant are:

  • An insufficient number of “good guy” bacteria — I do NOT mean traditional internet-lore good guy bacteria.
  • An insufficient number of specific enzymes being produced.

In every study we have had 99% of the items being statistically significant at the 99% level being deficiency, i.e. with a condition, the mean of those with the condition is less than the mean of the reference group. The 1% that show up as too high are usually at relatively low z-score and is a surprising good match to the number expected from the False Detection Rate. In other words, they were false significant.

This means that some refactoring is planned. The first one is to give an option to get suggestions ONLY from the undergrowth. The second one, revised the probiotic suggestions from using the genomic data from KEGG: Kyoto Encyclopedia of Genes and Genomes to match the pattern. This has just been done.

A common pattern that I have seen in the enzyme data has been that the mean of the study group was about 30% of the mean of the reference group. So, the shotgun approach is to flag in a sample all of the enzymes that are less than 30% of the refence group (using lab specific data). From the flagged enzymes, we look at all of the available probiotics as well as potential probiotics species and see which ones produces the most of the missing enzymes. My expectation is that often the best of these probiotics may “kick ass”, in the case of a ME/CFS person “Just started it and wow does this one give me crazy dreams and headaches. I even tried it in the morning and still the same result.  Did you just power though or any pointers to deal with the herx” 

PLACEHOLDER — VIDEO COMING

Special Studies: Anxiety

This is a common symptom for our times. This is reported often in samples, and thus being examined if it reaches our threshold for inclusion as defined in A new specialized selection of suggestions links. It does.

Anxiety is likely a part of a feedback look — anxiety chemicals modify the microbiome resulting in more anxiety triggering chemicals being released.

Study Populations:

We have 3 symptom annotations that could be included

  • Comorbid: High Anxiety [8.5 z-score x 45 samples]
  • DePaul University Fatigue Questionnaire : Anxiety/tension [6.5 z-score x 63 samples]
  • Condition: Generalized anxiety disorder [just 7 samples – so we exclude]

Taking the first two together we get 85 samples with an max z-score of 5.7

SymptomReferenceStudy
High Anxiety and Anxiety/tension111585
  • Bacteria Detected with z-score > 2.6: found 160 items, highest value was 5.7
  • Enzymes Detected with z-score > 2.6: found 290 items, highest value was 5.7
  • Compound Detected with z-score > 2.6: found No items

This hints that the enzymes being produced are likely more significant than the bacteria.

Interesting Significant Bacteria

All bacteria found significant (except 1) had too low levels. This is a common pattern found with these studies, it is not “bad bacteria bogie man bacteria” but an absence of “upstanding citizens bacteria”.

BacteriaReference MeanStudyZ-Score
Actinobacillus pleuropneumoniae (species)56215.7
Haemophilus parainfluenzae (species)16335435.5
Haemophilus (genus)16325455.4
Actinobacillus porcinus (species)183685.2
Pasteurellaceae (family)20017605.1
Pasteurellales (order)20017605.1

This is echoed in some studies

Interesting Enzymes

All enzymes found significant had too low levels.

EnzymeReference MeanStudy MeanZ-Score
HslU—HslV peptidase; (3.4.25.2)717732185.7
UDP-N-acetyl-alpha-D-glucosamine hydro-lyase (configuration-retaining; UDP-2-acetamido-2,6-dideoxy-alpha-D-xylo-hex-4-ulose-forming) (4.2.1.135)18965195.6
propanoate:CoA ligase (AMP-forming) (6.2.1.17)19216145.4
[RNA] 5′-hydroxy-ribonucleotide-3′-[RNA fragment]-lyase (cyclicizing; [RNA fragment]-3′- nucleoside-2′,3′-cyclophosphate-forming and hydrolysing) (4.6.1.19)15415345.3
(2S,3R)-3-hydroxybutane-1,2,3-tricarboxylate pyruvate-lyase (succinate-forming) (4.1.3.30)14744645.3
(S)-2-hydroxyglutarate:quinone oxidoreductase (1.1.5.13)22417105.3
propanoyl-CoA:oxaloacetate C-propanoyltransferase (thioester-hydrolysing, 1-carboxyethyl-forming) (2.3.3.5)15195005.3
CMP-N-acetyl-beta-neuraminate:beta-D-galactosyl-(1->4)-N-acetyl-beta-D-glucosaminyl-R (2->3)-N-acetyl-alpha-neuraminyltransferase (configuration-inverting) (2.4.99.6)16305495.3
4-hydroxybutanoate:NAD+ oxidoreductase (1.1.1.61)23706805.3
ATP:(Kdo)-lipid IVA 3-deoxy-alpha-D-manno-oct-2-ulopyranose 4-phosphotransferase (2.7.1.166)15355555.2
n/a (3.1.25.1)15735565.1
(S)-3-amino-2-methylpropanoate:2-oxoglutarate aminotransferase (2.6.1.22)24248765.1
5-aminopentanoate:2-oxoglutarate aminotransferase (2.6.1.48)24318835.1

Usually I leave the enzymes to nerds, but in this I decided to do a little exploration. I recall that Lactobacillus Casei has been shown to reduce anxiety in studies. So I looked up what enzymes are in it. In the above list of too low, I found the first one above, 3.4.25.2 in it. It opens up a possible model for picking probiotics a priori to address anxiety.

There was only one lactobacillus with more than one of these enzymes, Lactobacillus yamanashiensis – not available as a retail probiotic. There were three bifidobacterium which had more than one of these enzymes:

  • Bifidobacterium dentium (3)
  • Bifidobacterium thermophilum (3)
  • Bifidobacterium subtile (2)

Unfortunately, none are available as a retail probiotic. HOWEVER, there is one commercial probiotic that appears to have EIGHT (8) of these enzymes: Escherichia coli, available as Mutaflor (E.Coli Nisse 1917) or Symbioflor-2. So if you have anxiety, the theoretically best probiotic for you to take is an E.Coli one!!

A surprise!

I went to look at the a priori suggestions and was shocked/delighted that the only probiotic on the suggestion list was Symbioflor-2. Even with the complexities of 160 bacteria to adjust, somehow our enzymes calculations and this totally study based approach came to agreement! I repeat, I was shocked —

This selecting probiotics based on enzyme deficiency has an interesting scent and I will likely follow up on future posts.

SIBO Bacteria — nothing is certain…

A small number of possible ones are reported in the literature, but the quality of the results are suspect.

Tax RankTax NameShiftCitation Link
genusEnterococcus (NCBI:1350 )High   📚 PubMed
genusKlebsiella (NCBI:570 )High   📚 PubMed
genusPrevotella (NCBI:838 )High   📚 PubMed
genusSalmonella (NCBI:590 )High   📚 PubMed
genusStaphylococcus (NCBI:1279 )High   📚 PubMed
genusStreptococcus (NCBI:1301 )High   📚 PubMed
phylumFirmicutes (NCBI:1239 )Low   📚 PubMed
speciesAcinetobacter baumannii (NCBI:470 )High   📚 PubMed
speciesBifidobacterium longum (NCBI:216816 )Low   📚 PubMed
speciesEnterococcus faecalis (NCBI:1351 )High   📚 PubMed
speciesEnterococcus faecalis (NCBI:1351 )Low
speciesEnterococcus faecium (NCBI:1352 )High   📚 PubMed
speciesEscherichia coli (NCBI:562 )High   📚 PubMed
speciesEscherichia coli (NCBI:562 )High
speciesKlebsiella pneumoniae (NCBI:573 )High   📚 PubMed
speciesMethanobrevibacter smithii (NCBI:2173 )High   📚 PubMed
speciesMethanobrevibacter smithii (NCBI:2173 )High
speciesPseudomonas aeruginosa (NCBI:287 )High   📚 PubMed
From https://microbiomeprescription.com/library/PubMedCitation?CondId=67

The usual method of testing is from breath tests. From KEGG.JP we can get a list of bacteria that produces the compounds detected in the breath — the number is huge.

This page/video was suggested reading/viewing by a reader.

Question: How do you search on Microbiome Prescription for H2 etc?

On My Own Sample

Which Test? Is GI-MAP not enough?

To me, the more information about the microbiome that you have, the better it is to identify issues and build a treatment course. Below, the numbers will speak for themselves.

Second Class Tests

I deem these as 2nd class for several reasons: the number of bacteria reported is low (compare to others), the lab method requires the bacteria to be culturable (hence many bacteria will never be reported), they do no provide a suitable CSV file for upload to Microbiome Prescription (allowing 2nd opinions for treatment), they mechanism of measurement is not compatible to Microbiome Prescription (and most recent microbiome studies on the US National Library of Medicine)

Lab NameBacteriaReported
Bioscreen (cfu/gm)17
Biovis Microbiome Plus (cfu/g)40
DayTwo76
Diagnostic Solution GI-Map (cfu/gm)38
GanzImmun Diagnostic A6 (cfu/gm)76
GanzImmun Diagnostics AG Befundbericht25
Genova Gi Effects (cfu/g)28
Genova Parasitology (cfu/g)7
GI EcologiX (Invivo)55
GI360 Stool (UK)67
Gut Zoomer (vibrant-wellness)152
InVitaLab (cfu/gm)23
Kyber Kompakt (cfu/g)11
Medivere: Darm Mikrobiom Stuhltest (16s limited)16
Medivere: Darn Magen Diagnostik (16s Limited)16
Medivere: Gesundsheitscheck Darm (16s Limited)17
Metagenomics Stool (De Meirleir) (16s Limited)53
Smart Gut (ubiome 16s – Limited Taxonomy)23
Verisana (cfu/ml) aka (kbe/ml)11
Viome (No objective measures)29

First Class Tests

These are first class because: they allow an easy upload to Microbiome Prescription, use an appropriate measurement process, report based on bacteria DNA/RNA and not cultured count. The key words are 16s and shotgun analysis and not cultured.

There can be considerable cost differences between these labs (links to my first choices are below).

Lab/ProcessorLowTypicalHighUploads
AmericanGut7315621318
BiomeSight1157013051283
BiomeSightRdp27965686211
Nirvana / CosmosId643070533
es-xenogene1461288052254
Medivere5307219347
Microba5312215316
SequentiaBiotech16631346036
Ombre Labs18566722381066
uBiome (Out of Business)6249589813

Most Useful Lab?

Microbiome Prescription does some Fuzzy Logic Artificial Intelligence (FLAI) a.k.a. Dr. Artificial Intelligence. The accuracy is a function of the data available to him, that is the number of uploads from a specific lab. The newer lab BiomeSight is the winner here. This preference is made stronger because they will ship worldwide while #2, OmbreLabs (according to feedback) will only ship to the US. Microbiome Prescription mission is to support people worldwide .. so the handwriting on the statistics board is clear.

We have a growing number of special studies using BiomeSight data.

  1. 🗺️ Allergic Rhinitis (Hay Fever) ( 165 candidate bacteria) with highest z-score of 9.5
  2. 🗺️ Autism ( 192 candidate bacteria) with highest z-score of 8.3
  3. 🗺️ Bloating ( 126 candidate bacteria) with highest z-score of 5.4
  4. 🗺️ Brain Fog ( 93 candidate bacteria) with highest z-score of 5.2
  5. 🗺️ Chronic Fatigue Syndrome (CFS/ME) ( 170 candidate bacteria) with highest z-score of 6.6
  6. 🗺️ Cold Extremities ( 145 candidate bacteria) with highest z-score of 12.7
  7. 🗺️ COVID19 (Long Hauler) ( 218 candidate bacteria) with highest z-score of 12.6
  8. 🗺️ Easily irritated ( 248 candidate bacteria) with highest z-score of 9.9
  9. 🗺️ High Anxiety ( 170 candidate bacteria) with highest z-score of 5.9
  10. 🗺️ Histamine or Mast Cell issues ( 138 candidate bacteria) with highest z-score of 8.4
  11. 🗺️ Inflammatory bowel disease ( 245 candidate bacteria) with highest z-score of 12.5
  12. 🗺️ irritable bowel syndrome ( 150 candidate bacteria) with highest z-score of 6.7
  13. 🗺️ ME/CFS with IBS ( 193 candidate bacteria) with highest z-score of 8.4
  14. 🗺️ ME/CFS without IBS ( 251 candidate bacteria) with highest z-score of 7.3
  15. 🗺️ Post-exertional malaise ( 175 candidate bacteria) with highest z-score of 6
  16. 🗺️ Tinnitus (ringing in ear) ( 133 candidate bacteria) with highest z-score of 6.7
  17. 🗺️ Unrefreshed sleep ( 165 candidate bacteria) with highest z-score of 6.9

Social Media Questions

For other tests not covered above, see 16s Providers

On Sun Genomics

On Gut Zoomer

“Show me the beef!” This appears to be pure marketing hype — or more specifically, a substance for microbiome analysis that is high in pathogens!

Gut Zoomer [Est. 2016] – 170 -200 species (per their advertising) must be ordered thru a physician

https://www.vibrant-wellness.com/tests/gut-zoomer/

While Xerogene typically reports TEN TIMES as many!!! It is DEFINITELY NOT THE MOST COMPLETE, it is infact at the bottom of their first class competitors but the top of second class competitors.

What about Biohm

Based on their official View a sample report. they report on only 44 bacteria items.

Special Studies: Irritable Bowel Syndrome

We have IBS annotated in three different ways:

  • Autonomic Manifestations: irritable bowel syndrome
  • Official Diagnosis: Irritable Bowel Syndrome
  • Condition: ME/CFS with IBS

I ran each possible combination of the above and Official Diagnosis: Irritable Bowel Syndrome gave the strongest results and this will be reported here. When other things are combined, it is common for associations to become diffused/weaker. This is part of  A new specialized selection of suggestions links.

Study Populations:

SymptomReferenceStudy
Official Diagnosis: Irritable Bowel Syndrome112256
  • Bacteria Detected with z-score > 2.6: found 148 items, highest value was 6.7
  • Enzymes Detected with z-score > 2.6: found 218 items, highest value was 6.1
  • Compound Detected with z-score > 2.6: found No items

Interesting Significant Bacteria

The results are very striking — low bifidobacterium across the board at the top! Almost all of the bacteria associated are low.

BacteriaReference MeanStudyZ-Score
Bifidobacterium gallicum (species)37415876.7
Bifidobacterium kashiwanohense PV20-2 (strain)325626.6
Bifidobacterium catenulatum subsp. kashiwanohense (subspecies)315626.6
Thermosediminibacterales (order)49166.3
Lactiplantibacillus pentosus (species)123285
Escherichia (genus)598313974.9
Bifidobacterium cuniculi (species)81304.7
Bifidobacterium angulatum (species)184324.7

Interesting Enzymes

As is often seen with various symptoms/conditions, the associations are due to insufficiency and not surplus.

EnzymeReference MeanStudy MeanZ-Score
substrate,NADPH—hemoprotein reductase:oxygen oxidoreductase (RH-hydroxylating or -epoxidizing) (1.14.14.1)51196.1
NADPH:hemoprotein oxidoreductase (1.6.2.4)51196.1
propanoyl-CoA:oxaloacetate C-propanoyltransferase (thioester-hydrolysing, 1-carboxyethyl-forming) (2.3.3.5)14803816.1
(2S)-3-(4-hydroxyphenyl)-2-isocyanopropanoate,2-oxoglutarate:oxygen oxidoreductase (decarboxylating) (1.14.20.10)131266
(2S)-3-(4-hydroxyphenyl)-2-isocyanopropanoate,2-oxoglutarate:oxygen oxidoreductase (1.14.20.9)131266
(2S,3R)-3-hydroxybutane-1,2,3-tricarboxylate pyruvate-lyase (succinate-forming) (4.1.3.30)14323755.9
ATP:amicoumacin A 2-phosphotransferase (2.7.1.230)41185.7
(E)-4-(trimethylammonio)but-2-enoyl-CoA:L-carnitine CoA-transferase (2.8.3.21)13543885.7
gamma-butyrobetainyl-CoA:electron-transfer flavoprotein 2,3-oxidoreductase (1.3.8.13)13653905.7
L-carnitine:CoA ligase (AMP-forming) (6.2.1.48)13394435.4
hydrogen-sulfide:ferredoxin oxidoreductase (1.8.7.1)991246835.3
NADPH:acceptor oxidoreductase (1.6.99.1)568422845.3
(2S,3S)-2-hydroxybutane-1,2,3-tricarboxylate hydro-lyase [(Z)-but-2-ene-1,2,3-tricarboxylate-forming] (4.2.1.79)12844075.2
S-methyl-5′-thioadenosine:phosphate S-methyl-5-thio-alpha-D-ribosyl-transferase (2.4.2.28)363816685.1
acyl-CoA,ferrocytochrome b5:oxygen oxidoreductase (6,7 cis-dehydrogenating) (1.14.19.3)10622845
butanoyl-CoA:acetoacetate CoA-transferase (2.8.3.9)389222295
L-carnitinyl-CoA hydro-lyase [(E)-4-(trimethylammonio)but-2-enoyl-CoA-forming] (4.2.1.149)14223625

Bottom Line

Looking at the bacteria, the probiotic solution would be:

  • Bifidobacterium (a wide variety)
  • E. Coli Probiotic (Mutaflor or Symbioflor-2)

Lactobacillus are HOSTILE to E.Coli, so Lactobacillus probiotics should likely be avoided.

Looking at some of the recommendations, we see Human milk oligosaccharides (prebiotic, Holigos, Stachyose) which are known to greatly encourage bifidobacterium. We also see saccharomyces boulardii which is also known to increase Bifidobacterium [2020].

As always, using your own 16s samples would produce the best suggestions.

https://microbiomeprescription.com/Library/CitizenScience

Suggestions Conflicts

A reader messaged me the following concern

hi ken, what would you do if there’s a massive contradiction between most of the suggestions generated by the general consensus and the suggestions generated by your new “from special studies” biomesight algorithm

The first thing that I want to point out is the warning on that page.

The main issue you may be seeing is in the selection of bacteria. With the regular selection, you focus on extreme values, i.e. top 10%, outside of standard lab ranges, outside of reference ranges from Jason Hawrelak and others. The amount outside of the reference range is used to give a weight to each bacteria for the importance of shifting. Different algorithms are used with different approaches (we do not know what the ideal one is).

With the special studies, we up-ended the algorithm. We picked the bacteria based on a simple “if the amount is above or below the reference norm+/- twice the standard deviation of the mean for the reference population and then use the z-score as the weight (the statistical significance for this bacteria)”.

This change means that a bacteria that is at the 70%ile may be included in the selection (which is very unlikely with the the first methods), and this bacteria could have a very high weight (which is based on statistical significance and NOT the difference from a mile post). A Bacteria at the 99%ile will be totally ignored if it is not statistically significant for the condition.

Statistically, I prefer the special studies approach because we are using the statistical significance of the bacteria for the significance/weight for suggestions instead of the naïve assuming that being high or low is the cause.

Bottom Line

  • We pick bacteria based on statistical significance for a specific condition and not whether they are high or low in general
  • We give the bacteria a weight based on statistical significance for a specific condition and not the difference from a bound.

In theory, with the identical same bacteria and counts selected for two different conditions, you will get different suggestions because the weight assigned will be different since the weight is based on the statistical significance for the condition.

I well understand the confusion of some, the model being used is getting more advanced and handling more complexities.

Special Studies: Cold Extremities

This is a common symptom for both ME/CFS and Long COVID. This is reported often in samples, and thus being examined if it reaches our threshold for inclusion as defined in A new specialized selection of suggestions links. It does.

My default view is that this symptom is likely due to vesicular constriction/inflammation or “stick blood” (which contains many coagulation possibilities) resulting in low warming blood flow.

Study Populations:

SymptomReferenceStudy
Sleep: Unrefreshed Sleep108778
  • Bacteria Detected with z-score > 2.6: found 145 items, highest value was 12.6
  • Enzymes Detected with z-score > 2.6: found 170 items, highest value was 5.3
  • Compound Detected with z-score > 2.6: found No items

The highest z-scores above are more than most other symptoms. This indicates that the causes are more homogeneous bacteria shifts.

For those with uploaded microbiome

Interesting Significant Bacteria

All bacteria found significant had too low levels. The dominant genus is Prevotella with the P.Copri being usually 90% of the genus count in our reference, but only 14% of the genus count with Cold Extremities indicating that the shift is very species specific. Looking at Prevotella copri in isolation, we see the best documented ways of increasing it are: mediterranean diet, resveratrol (grape seed/polyphenols/red wine), berberine, navy bean,  Conjugated Linoleic Acid and linseed(flaxseed).

It is interesting to note that some of these are known to improve coagulation or reduces inflammation

BacteriaReference MeanStudyZ-Score
Prevotella copri (species)66900309012.6
Prevotella (genus)74698212937.6
Prevotellaceae (family)82172366086.1
Bifidobacterium kashiwanohense PV20-2 (strain)326785.9
Bifidobacterium catenulatum subsp. kashiwanohense (subspecies)317785.8
Sporolactobacillaceae (family)173585.7
Thermosediminibacterales (order)48185.7
Sporolactobacillus (genus)174605.5
Sporolactobacillus putidus (species)174605.5
Bifidobacterium gallicum (species)37847835.5
Lactiplantibacillus pentosus (species)123225.3
Phocaeicola coprocola (species)75846635

Interesting Enzymes

Most (99%) enzymes found significant had too low levels.

EnzymeReference MeanStudy MeanZ-Score
n/a (3.4.24.20)40165.3
2-acetylphloroglucinol C-acetyltransferase (2.3.1.272)182585.1
ATP:L-threonine O3-phosphotransferase (2.7.1.177)24165295

Bottom Line

In this study, one species shouts out as the cause by not being there Prevotella copri. I would really emphasis the items listed above as encouraging its growth.

It may be available “soon” as a probiotic, “The gut bacterium Prevotella copri (P. copri) has been shown to lower blood glucose levels in mice as well as in healthy humans, and is a promising candidate for a next generation probiotic aiming at prevention or treatment of obesity and type 2 diabetes.” [2021]

This article is an interesting read: The Strange Case of Prevotella copri: Dr. Jekyll or Mr. Hyde? Especially for those who think all bacteria can be classified as either good or bad.

In terms of retail probiotics suggested, the E.Coli probiotic, Symbioflor-2 is suggested.

To obtain a priori suggestions, go to https://microbiomeprescription.com/Library/CitizenScience

The a priori suggestions, shown below, contains the items for P. Copri cited above.

Social Feedback
From Facebook

Special Studies: Unrefreshing Sleep

This is a common symptom for both ME/CFS and Long COVID. This is reported often in samples, and thus being examined if it reaches our threshold for inclusion as defined in A new specialized selection of suggestions links. It does, but the degree of association (z-scores) are lower than prior special studies despite having a larger study population.

Study Populations:

SymptomReferenceStudy
Sleep: Unrefreshed Sleep1041107
  • Bacteria Detected with z-score > 2.6: found 139 items, highest value was 5.1
  • Enzymes Detected with z-score > 2.6: found 208 items, highest value was 5.2
  • Compound Detected with z-score > 2.6: found 1 items, highest value was 2.8 – effectively zero when false detection rate is considered.

The highest z-scores above are less than other symptoms despite larger sample size. This indicates that the causes are more diverse and thus less homogeneous bacteria shifts.

Location of this custom filter

Interesting Significant Bacteria

All bacteria found significant had too low levels.

Low Veillonella is reported in some studies associated with sleep issues with it’s consumption of lactic acid(lactate) being cited as a possible factor:

BacteriaReference MeanStudyZ-Score
Veillonella (genus)406322825.1
Actinobacillus (genus)3531334.9
Bifidobacterium catenulatum subsp. kashiwanohense (subspecies)3211004.8
Clostridium cellulovorans (species)40174.8
Bifidobacterium kashiwanohense PV20-2 (strain)3181004.8
Actinobacillus porcinus (species)184734.6
Thiobacillus thiophilus (species)85264.6

Interesting Enzymes

All enzymes found significant had too low levels.

EnzymeReference MeanLong COVID MeanZ-Score
(1->4)-alpha-D-galacturonan lyase (4.2.2.2)245310745.2
2-acetylphloroglucinol C-acetyltransferase (2.3.1.272)185614.6
phylloquinone:disulfide oxidoreductase (1.17.4.4)28154.6
[RNA] 5′-hydroxy-ribonucleotide-3′-[RNA fragment]-lyase (cyclicizing; [RNA fragment]-3′- nucleoside-2′,3′-cyclophosphate-forming and hydrolysing) (4.6.1.19)15966344.5
CMP-N-acetyl-beta-neuraminate:beta-D-galactosyl-(1->4)-N-acetyl-beta-D-glucosaminyl-R (2->3)-N-acetyl-alpha-neuraminyltransferase (configuration-inverting) (2.4.99.6)16826594.5
propane-1,3-diol:NAD+ 1-oxidoreductase (1.1.1.202)8111064.5
ATP:(Kdo)-lipid IVA 3-deoxy-alpha-D-manno-oct-2-ulopyranose 4-phosphotransferase (2.7.1.166)15916454.5

Bottom Line

The key take away is that lactate/lactic acid levels appears to be a significant contributor. The reason that it is high is the lack of lactate consumers. Removal of lactate producing probiotics from supplements appears to be a logical first step (i.e. Lactobacillus), followed by taking Vitamin B1 shortly before bed time.

For suggestions on lowering it, see this old post of mine in the ME/CFS context.

The issue may not be lactate by itself, but by the form of lactate (d-lactate –> bad)

Results and conclusion: Administration of L-lactate does not influence sleep-wake cycle of experimental animals. At the same time, its artificial optical analog D-lactate induces the significant (as compared to the control) decrease in wake (34.8% to 26.5%) and increase in slow wave sleep (57.4% to 69.2%). It has been suggested that D-lactate may be the antagonist of one or several L-lactate receptors.

[D-lactate as a novel somnogenic factor?] [2020]

“Thiamine (Vitamin B1) replenishment at intravenous doses of 100 mg every 12 h resolved lactic acidosis and improved the clinical condition in 3 patients.” [1997]

“Lactomin[300 mg Lactobacillus acidophilus, 300 mg Bifidobacterium longum] was discontinued, and she was treated with sodium bicarbonate and oral antibiotics. The probiotics the patient had taken were likely the cause of D-lactic acidosis ” [2010]

Phospholipids, the Microbiome and Mr. Hughes

For almost a decade I have suspected that there was an interaction between the microbiome and Antiphospholipid syndrome (APS) also known as Hughes Syndrome (after the MD, see below). This is also called  “sticky blood syndrome” [HealthLine]. For some researchers, it is deemed to be a significant contributor to fatigue in Myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) [1999 D. Berg] and likely also applies to Long COVID. My own singleton experience seems to confirm it for myself.

A reader asked about phospholipids on Facebook today, so I revisited available literature

This article by Graham R.V. Hughes, MD, FRCP (the discoverer) in 2016 is well worth reading.

For me, APS/Hughes syndrome is very much a neurological condition. Brain function does seem to be especially targeted—the more APS patients one sees, the wider and wider the neuropsychiatric ripples spread.

APS: What Rheumatologists Should Know about Hughes Syndrome • By Graham R.V. Hughes, MD, FRCP

Of course, running off the experience of just one, or even a few people, is not the best practice. Testimonials suck because of rose color glasses, fake testimonials, mainly positive responders report, and placebo effects. So what does the literature state. First there is some literature that are general discussions without the type of detail that I would love to see:

Then we come to this article: Phosphatidylglycerols are induced by gut dysbiosis and inflammation, and favorably modulate adipose tissue remodeling in obesity [2019] which uses one of my favorite information source, the Kyoto Encyclopedia of Genes and Genomes. “We found that PGs were positively associated with microbiomes enriched with endotoxin-synthesis genes and associated with markers of inflammation.”

Digging further we find:

 Bacteroides thetaiotaomicron, Actinomyces massiliensis, Pseudopropionibacterium propionicum, Corynebacterium amycolatum, Ruminococcus gnavus and Roseburia intestinalis[2021] lead to the formation of pathogenic T‑cell and autoantibody responses via the cross-reactivity with autoantigens (Ro60, dsDNA and ß2 glycoprotein I). 

The role of the microbiome in lupus and antiphospholipid syndrome [2020]

M. pneumoniae and Streptococcus spp. infections, which are among the most prevalent bacterial infections in children and young adults, were linked to the occurrence of aPL. …. an anaerobic bacterium Fusobacterium necrophorum, although a variety of other bacteria such as streptococci, staphylococci, and enterococci may be also responsible…. a specific change in the gut microbial composition in APS patients. Particularly, a decrease of bacteria belonging to the genus Bilophila and overgrowth of bacteria of the Slackia genus were shown…  enrichment by Slackia spp. and by the lower abundance of butyrate-producing Butyricimonas 

Environmental Triggers of Autoreactive Responses: Induction of Antiphospholipid Antibody Formation [2019]

More discussion of mechanism is in The Role of the Gut Microbiota in the Pathogenesis of Antiphospholipid Syndrome [2015]

Bottom Line

APS only requires one of the bacteria above to trigger it. In terms of using Microbiome Prescription, I would look at Bilophila and Butyricimonas – if below 50%ile, hand pick it, then look at Slackia, if above 50%ile then hand pick it. Check the other bacteria cited above, and if any are over 75%ile, hand pick those. “It only takes one rotten apple to spoil the barrel” seems to apply here.

I have added APS to my PubMed reference list:

Personal Observations

I checked my samples from my last ME/CFS flare and found that Bacteroides thetaiotaomicron went from 73%ile on first sample after onset, to 96%ile on second sample, down to 79%ile, then 70%ile then 20%ile a few months later with recovery and returning to work. The key triggering bacteria will likely be different for each person but you at least have a candidate list to work from.