Ascribing a condition to one (or even a few) bacteria may be very wrong

An illness to a single bacteria is a standard medical practice. Helicobacter pylori for ulcer, or one of several E. Coli (E. coli O157:H7 and other STECs like E. coli O145 and E. coli O121:H19) for food poisoning.

If we expand our eyes and ask “What if the condition is caused by over or under production of some metabolite?”

For CFS over production of d-lactic acid has been proposed, see this post

Let us assume that this is part of it. So which bacteria produces it?

Our old friend DataPunk lists:

PRODUCED AS ENDPRODUCT BY: 

  1. Aerococcus
  2. Allobaculum
  3. Anaerostipes
  4. Atopobium
  5. Bacillus coagulans
  6. Bifidobacterium
  7. Cardiobacteriales
  8. Cardiobacterium
  9. Carnobacterium
  10. Catenibacterium
  11. Enterococcus
  12. Faecalibaculum
  13. Faecalibaculum rodentium
  14. Gemella
  15. Haemophilus
  16. Holdemania
  17. Lachnobacterium
  18. Lactobacillales
  19. Lactobacillus
  20. Lactobacillus acidophilus
  21. Lactobacillus casei
  22. Lactobacillus delbrueckii
  23. Lactobacillus helveticus
  24. Lactobacillus rhamnosus
  25. Lactococcus
  26. Leptotrichia
  27. Leuconostoc
  28. Microbacterium
  29. Moryella
  30. Oenococcus
  31. Pediococcus
  32. Rothia
  33. Ruminococcus faecis
  34. Scardovia
  35. Serratia marcescens
  36. Streptococcus
  37. Tetragenococcus
  38. Vagococcus

Let us suppose that we also have low or no butyrate – again datapunk lists:

PRODUCED AS ENDPRODUCT BY:

 

  1. Agathobacter rectalis
  2. Allobaculum
  3. Anaerostipes
  4. Anaerostipes hadrus
  5. Anaerotruncus
  6. Butyricicoccus
  7. Butyricicoccus pullicaecorum
  8. Butyricicoccus sp. ORNL_6EZ5-Gt_4_Pl1-35
  9. Butyricicoccus sp. ORNL_6EZ5-Gt_6_Pl2-147
  10. Butyricicoccus sp. ORNL_V42_E05
  11. Butyricicoccus sp. ORNL_W42_C10
  12. Butyricimonas
  13. Catenibacterium
  14. Christensenella
  15. Cloacibacillus
  16. Cloacibacillus porcorum
  17. Clostridia
  18. Clostridiales
  19. Clostridium
  20. Clostridium butyricum
  21. Coprococcus
  22. Defluviitalea
  23. Eubacteriaceae
  24. Eubacterium limosum
  25. Eubacterium oxidoreducens
  26. Faecalibacterium
  27. Faecalibacterium prausnitzii
  28. Flavonifractor
  29. Flavonifractor plautii
  30. Fusibacter
  31. Fusobacterium
  32. Lachnobacterium
  33. Lachnospiraceae
  34. Moryella
  35. Moryella indoligenes
  36. Oscillospira
  37. Peptoniphilus
  38. Roseburia
  39. Roseburia faecis
  40. Roseburia hominis
  41. Roseburia intestinalis
  42. Roseburia inulinivorans
  43. Ruminococcus torques
  44. Subdoligranulum

So how many possible “too much and not enough” pairs are there if we just pick just one from each list…  38 x 44 = 1,672!!! 

My current gut feeling is that taking the % of each taxonomy in each group –> producing a fuzzy measure of a 0.23% Lactic acid and 0.01% butyrate, may reveal clearer patterns for symptoms than the individual bacteria taxonomy. This is a fuzzy measure because the production of lactic acid and butyrate vary greatly from genus to genus.

Other options for stepping away from pine needles of individual bacteria to looking at tree as a whole include KEGG pathways,

KEGG pathways produce some tentative correlation for D-Arginine and D-ornithine metabolism in CFS/IBS in my earlier post.

Bottom Line

The same change of metabolites in the body can happen in many many different ways. Looking at the metabolites (even with fuzzy ‘punted’ data base on what is available) may produce better predictive results than just looking at individual bacteria.