Saliva Genome 2- Microbiome

As I mentioned in a previous post: https://www.dhirubhai.net/posts/jeffrosenfeld_saliva-whole-genome-post-1-i-just-got-my-activity-7170876117258616832-l_6q/

I got free saliva whole genome sequencing (WGS) from Psomagen on a Complete Genomics machine.? Since the DNA was collected from saliva, I was curious about how much non-human DNA would be in the samples as well as the amount of my microbiome that would be present.? The oral microbiome has been well studied, so there are lots of reference samples.

?

In order to analyze the microbiome, I reached out to my good friend (and AGBT - Advances in Genome Biology and Technology buddy) Manoj Dadlani who leads CosmosID .? Cosmos is a leader in microbiome and they provide a full suite of wet-lab and bioinformatics solutions to analyze all types of microbiome samples.? Manoj Dadlani generously gave me some free credits on the COSMOS ID platform that I used for analysis.? In addition, Britton Strickland, Ph.D. gave me some assistance on the analysis and interpretation.

?

In contrast to many other WGS providers, Psomagen allows for download of the full FASTQ and BAM files that were produced.? For my genome, there were 700 million reads providing over 30X coverage of the genome.?

As a first check, I looked at the standard FASTQC metrics:

These show uniformly high quality and testify to the Complete Genomics sequencing fidelity.?

?

I was curious whether there would be enough microbiome to analyze, but a quick look showed that I had 18M bacterial reads in my sample.? Britton confirmed that this is more than sufficient for a microbial analysis.

I started analysis at the strain level.? This view showed a massively complicated circular tree showing hundreds of strains.

This was obviously too high a level of detail and is difficult to interpret.? The strain-level information may be too fine grained. Secondly, our understanding of the oral microbiome at the strain level is incomplete, so no real conclusions could be drawn from this view.?

? Zooming out to genus level gave me a much more useful picture.? Most of my oral microbiome consists of Neisseria, Prevotella and Veillonella which are normal components of a healthy mouth and not anything of concern.?

?

When I looked at fungi, the top hit was Melampsora pinitorqua. This taxa has not been regularly seen in humans and it is a plant parasite.? Most commonly, it is found to cause pie twisting rust.

?I would have dismissed finding as being spurious noise, but it was supported by 800k reads.? Potentially, this Melampsora was present in my mouth from the local Orlando tap water. It would take some further investigation to determine if this is a real hit or something like when we found bottlenose dolphin DNA in the New York subway ( Christopher Mason )

?

In conclusion, the human WGS kit did a great job of profiling my microbiome.

?

Fortunately, nothing worryingly pathogenic was found.?Thank you again to Manoj Dadlani , Britton Strickland, Ph.D. and the whole CosmosID team.

?

Have you done an oral genome or microbiome?

What have you found that was useful or worrying?

For those that are curious, the CosmosID tool is based on their own k-mer based microbiome algorithms

?