Bioreactor virome metagenomics sequencing using DNA spike-ins

Abstract

With the emergence of Next Generation Sequencing, major advances were made with regard to identifying viruses in natural environments. However, bioinformatical research on viruses is still limited because of the low amounts of viral DNA that can be obtained for analysis. To overcome this limitation, DNA is often amplified with multiple displacement amplification (MDA), which may cause an unavoidable bias. Here, we describe a case study in which the virome of a bioreactor is sequenced using Ion Torrent technology. DNA-spiking of samples is compared with MDA-amplified samples. DNA for spiking was obtained by amplifying a bacterial 16S rRNA gene. After sequencing, the 16S rRNA gene reads were removed by mapping to the Silva database. Three samples were tested, a whole genome from Enterobacteria P1 Phage and two viral metagenomes from an infected bioreactor. For one sample, the new DNA-spiking protocol was compared with the MDA technique. When MDA was applied, the overall GC content of the reads showed a bias towards lower GC%, indicating a change in composition of the DNA sample. Assemblies using all available reads from both MDA and the DNA-spiked samples resulted in six viral genomes. All six genomes could be almost completely retrieved (97.9%–100%) when mapping the reads from the DNA-spiked sample to those six genomes. In contrast, 6.3%–77.7% of three viral genomes was covered by reads obtained using the MDA amplification method and only three were nearly fully covered (97.4%–100%). This case study shows that DNA-spiking could be a simple and inexpensive alternative with very low bias for sequencing of metagenomes for which low amounts of DNA are available.

Soil viruses are underexplored players in ecosystem carbon processing

“…First, resuspending viruses from soils is a challenge due to their adsorption to the soil matrix (43). Second, yields of viral DNA are often very low (due to both low input biomass and potentially low extraction efficiency), requiring amplification; this leads to biases (53, 56–61) or poor assembly and few viral contigs (described further in 125). Third, viral contig identification requires a reference database, yet soil viruses are underrepresented in current databases; for example, a majority (85%) of our sequence space was unknown.…”

“…Due to the requirement of habitat-specific resuspension optimization, soil viromics is in its early stages. In addition, because particle yields are typically low, most soil virome studies have amplified extracted viral DNA using multiple displacement amplification, which renders the datasets both stochastically and systematically biased and non-quantitative (53, 56–61). The few polar soil viromes have been from Antarctic soils, and further demonstrated the genetic novelty of this gene pool, while suggesting resident viral communities were dominated by tailed-viruses, had high habitat specificity, and were structured by pH (62–64).…”

Soil Viruses Are Underexplored Players in Ecosystem Carbon Processing

“…Due to the requirement of habitat-specific resuspension optimization, soil viromics is in its early stages. In addition, because particle yields are typically low, most soil virome studies have amplified extracted viral DNA using multiple displacement amplification, which renders the data sets both stochastically and systematically biased and nonquantitative ( 54 , 58 – 63 ). The few polar soil viromes have been from Antarctic soils and further demonstrated the genetic novelty of this gene pool while suggesting that resident viral communities were dominated by tailed viruses, had high habitat specificity, and were structured by pH ( 51 , 64 , 65 ).…”