The first snowfall of the year is always so magical. Let’s just say it’s somewhat less magical as we approach the middle of April.😂

The first snowfall of the year is always so magical. Let’s just say it’s somewhat less magical as we approach the middle of April.😂

Finally, we put all the major findings regarding host-microbe and microbe-microbe interactions in a nice summary made by Clara. This shows the major functions and species for pH, CAMP, Fe, collagenase and amine metabolism.

Investigating only BV samples we found a robust split between two BV subtypes which we cunningly named BV1 and BV2. BV1 had higher expression of glycan and protein degradation genes, biotin metabolism and more oxidative phosphorylation expression than BV2.

New findings, accounting for scale: H and BV species have very different approaches to evade host-encoded cationic antimicrobial peptides (CAMPs; Fig 1D). We found many unintuitively-named and non-assigned KEGG functions in this analysis--combing through data labelled ‘unknown’ pays off sometimes!

Comparing H and BV we first found the expected hallmarks: butanoate metabolism and sialidase activity were higher in BV metatranscriptomes, while peptidoglycan biosynthesis was higher in healthy samples dominated by Gram-positive lactobacilli.

Now part 2: the functional analysis for our vaginal meta-transcriptome paper https://microbiomejournal.biomedcentral.com/articles/10.1186/s40168-024-01992-w. This is complicated because of the difference in species composition, abundance and gene content in each sample. More in the methods tomorrow.

6) Gardnerella as a genus is very diffuse and spans the entire spectrum of BV types. Looking at a biplot of only the BV samples, we found that different Gardnerella species occupied discrete areas (Sup Fig 5) but high sequence similarity made separating most of their transcripts very difficult

Fig 2B is more interesting and shows a compositional biplot of the transcripts mapped to the VIRGO database (colours = presumed species of origin). Healthy (H) vs. bacterial vaginosis (BV) samples separate cleanly on the first principle component. This is expected and is nice to see!

So let's dive into the descriptive data: Fig 2A shows the major species and how they associate with community state types. This recapitulates almost every vaginal microbiome paper: ‘healthy’ microbiomes are generally dominated by lactobacilli, mixed anaerobes predominate in BV.

The journey has taken over 10 years from sample collection to publication because existing tools were not designed to analyze metatranscriptomes and so new methods needed to be devised. We ended up conducting a meta-analysis of three different datasets to ensure our conclusions were robust.

mine

oh the horror stories that come out of public datasets

Bluesky now has over 10 million users, and I was #597,896!

Feeling like adding some content that’s not doom and gloom. So I will be posting some pictures I took on a recent scuba holiday to Turks and Caicos. First up this curious cutie