Microbiome Yarns: microbiome of the built environment, paranormal microbiology, and the power of single cell genomics1,2,3,4

Ontology-Aware Deep Learning Enables Ultrafast, Accurate and Interpretable Source Tracking among Sub-Million Microbial Community Samples from Hundreds of Niches

“…Our analysis results by using ONN4MST have shown that microbial communities on these surfaces mostly originated from humans ( Fig. 4a ), largely in agreement with the original analyses of Lax et al 21 using SourceTracker, and differs slightly from the results of Shenhav et al 9 These results were reasonable considering the strong influence of skin microbial communities on indoor house surfaces 22 , while they have again emphasized the challenge of source disambiguation for methods that do not consider ontology structure of the biomes. That is, treating each individual sample as an independent potential source would make differentiation of tiny sample differences among ontologically-close biomes impossible, thus underestimating the contributions of known sources at higher layers.…”

Ontology-aware deep learning enables ultrafast and interpretable source tracking among sub-million microbial community samples from hundreds of niches

“…4 a), largely in agreement with the original analyses of Lax et al [ 24 ] using SourceTracker, and differs slightly from the results of Shenhav et al . [ 12 ] These results were reasonable considering the strong influence of skin microbial communities on indoor house surfaces [ 30 ], while they have again emphasized the challenge of source disambiguation for methods that do not consider ontology structure of the biomes. That is, treating each individual sample as an independent potential source would make differentiation of tiny sample differences among ontologically-close biomes impossible, thus underestimating the contributions of known sources at higher layers.…”

Microbiome Yarns: bacterial predators, tissue tropism and molecular decoys

“…J Infect Dis 199: 1479-1487; Banerjee, A., Kim, B.J., Carmona, E.M., et al (2011) Bacterial pili exploit integrin machinery to promote immune activation and efficient blood-brain barrier penetration. Nature Comm 2: 462 9 see Timmis K. and Jebok F. (2018) Microbiome Yarns: human milk oligosaccharides, Bifidobacterium and immunopowergames. Microb Biotechnol 11: 437-441.…”

“…Jilly: Absolutely! At the mo, I am having such fun collaborating with Vic Torde of the Lorenzo von Syntech High Security Institute for Artificial Life in Madrid 18 on a variety of projects, mostly to do with the production of synthetic decoys 19 , molecules that mimic the surface receptors of our cellsthe cellular doorsonto which pathogens dock in order to enter and cause mischief in them. When delivered by an appropriate route in an infected person, molecular decoys have the potential to neutralise a pathogen, which attaches to the decoys instead of our cells.…”