This is a “scribe notes” post. I am working on implementing odds ratio as a forecaster for ME/CFS and encountered some issues. In my work experience, this means taking a significant step backwards to look at the data better. I will look at the bacteria with a high frequency of being reported in tests first. The genus of greatest interest from preliminary work are:

• Blautia
• Faecalibacterium
• Clostridium
• Lachnospira
• Streptococcus
• Collinsella
• Anaerostipes
• Parabacteroides
• Anaerotruncus
• Bifidobacterium
• Pseudobutyrivibrio

The data is from BiomeSight samples. The red line is what would be expected with no influence. The population percentile included those with ME/CFS. Removing them would increase the differences more (I went lazy).

Charts

Blautia

Lower levels are clearly significant.

Faecalibacterium

Lower levels are some significance.

Clostridium

Lower levels are clearly significant.

Lachnospira

Higher levels are some significance for low values.

Streptococcus

Does not appear to have much significance

Collinsella

Higher levels are some significance for below average values

Anaerostipes

Lower levels are significance for all values.

Parabacteroides

Does not appear to have much significance

Anaerotruncus
Lower levels are significance for all values.

Bifidobacterium

Higher levels are significance for all values.

Pseudobutyrivibrio

Higher levels are significance for all values.

Next Steps

We can visually see how bacteria is shifted for these genus. The challenge is converting it to a formula to forecast that cross validates. Stay tune.