mNGS Nanopore: Initial Pipeline Validation

 

Overview

The CZ ID mNGS Nanopore pipeline was developed in collaboration with the bioinformatics team from Oxford Nanopore Technologies (ONT). Our team established baseline performance on key pipeline outputs and used these to refine the initial version of the pipeline prior to beta release. First, we identified and acquired five key datasets detailed below. Then, we evaluated pipeline performance on these datasets to document bugs and fix issues before launching the pipeline.

Below we highlight key results from five datasets that were used to validate the mNGS Nanopore pipeline, including:

Viral spike-in to human cell line 

Mock microbial community 

Mosquito samples

Human microbiome (fecal sample)

Environmental microbiome (hot springs)

---

Dataset: Viral spike-in to human cell line (24 hr infection)

Validation objective:

Detect a known spiked-in pathogen at varying concentrations or multiplicity of infection (MOI)

Sample background:

Pathogen: Human Betacoronavirus (HCoV) OC43 

Host: Human cell line

Nucleic acid type: RNA

Sequencer: GridION

Pore: R9.4.1

Basecaller: Guppy v5

Source: Andreas Puschnik (CZ Biohub Rapid Response Team), resequenced by ONT team

Key Result:

CZ ID pipeline detects betacoronavirus at six different MOI, including the lowest level of infection.

Sample	MOI	% HCoV OC43  (of total bp)
Human-1	1	25
Human-2	.1	1
Human-3	.01	3
Human-4	.001	0.3
Human-5	.0001	0.03
Human-6	0 (no virus)	0

---

Dataset: Mock microbial community standard

Objective:

Detect organisms at known relative abundances using the ZymoBIOMICS Microbial Community Standard. This standard is a well-defined mixed microbial community that can be used to optimize workflows or as a standard for quality control. Notably, this ZymoBIOMICS standard was evaluated in a previous study by Nicholls et al. 2019 after sequencing the standard through nanopore technology.

Sample Background:

Nucleic acid type: DNA

Sequencer: GridION

Pore: R9.4.1

Basecaller: Guppy v6

Source: Miten Jain (Northeastern University)

Theoretical composition based on genomic DNA:

3 easy-to-lyse Gram-negative bacteria

Listeria monocytogenes - 12%

Pseudomonas aeruginosa - 12%

Bacillus subtilis - 12%

5 tough-to-lyse Gram-positive bacteria

Escherichia coli - 12%

Salmonella enterica - 12%

Lactobacillus fermentum - 12%

Enterococcus faecalis - 12%

Staphylococcus aureus - 12%

2 tough-to-lyse yeasts

Saccharomyces cerevisiae - 2%

Cryptococcus neoformans - 2%

Key Result:

CZ ID pipeline detects all known organisms in the standard. Note that variation from the predicted standard community composition could be related to wet lab protocols conducted by the authors of the original study.

Relative abundances for organisms found in the ZymoBIOMICS community standard based on predicted abundances (Theoretical) and those detected through the CZ ID Nanopore pipeline and a study published by Nicholls et al. 2019.

---

Dataset: Mosquito samples previously sequenced using short-reads

Objective:

Detect viruses in complex samples from a non-human host and compare results from the CZ ID Nanopore pipeline to an Illumina (short-reads) run analyzed through the CZ ID Illumina pipeline.

Sample background:

Nucleic acid type: RNA

Sequencer: GridION

Pore: R9.4.1

Basecaller: Guppy v5

Source: Amy Kistler (CZ Biohub), resequenced by ONT team, Single mosquito metatranscriptomics identifies vectors, emerging pathogens and reservoirs in one assay | eLife

Key Result:

CZ ID Nanopore pipeline detects all previously characterized mosquito viruses detected through the Illumina pipeline. The table below summarizes detected viruses.

Mosquito	Previously Characterized Virus
Mosquito-1	Bunyavirus sp.
Mosquito-2	Culex narnavirus 1
Mosquito-3	Marma virus
	Culex bunya-like virus
	Wuhan Mosquito Virus 6
Mosquito-4	Culex flavivirus
	Culex bunya-like virus
	Marma virus
Mosquito-5	Hubei mosquito virus 2

---

Dataset: Human microbiome (fecal sample)

Objective:

Characterize complex sample from human host

Sample background:

Sample type: Human fecal sample

Nucleic acid type: DNA

Sequencer: MinION

Source: Brian Yu (from Biohub)

Key Result:

CZ ID Nanopore pipeline detected human fecal microbiota including bacteria and viruses.

Tree view from mNGS Nanopore sample report for human fecal sample.

---

Dataset: Environmental microbiome

Objective:

Characterize complex environmental sample

Sample background:

Sample location: Yellowstone National Park - Octopus Spring

Nucleic acid type: DNA

Sequencer: MinION

Source: Brian Yu (from Biohub)

Key Result:

The CZ ID Nanopore pipeline identified species that have been reported from the Octopus Spring in Yellowstone National Park, with the exception of E. coli. The unexpected detection of E. coli highlights the need to run negative controls alongside samples given that this bacterial species is a known laboratory contaminant.

Tree view from mNGS Nanopore sample report for sample collected at Octopus Spring.