DNA samples selected for long read sequencing (LRS) are routinely required to be 'pure' with high DNA concentration. Hence the usefulness of samples with substandard DNA quality for LRS is unknown. We aim to perform de-novo assembly of Adenovirus sequenced from non-human primate (NHP) faeces using the Oxford Nanopore technologies (ONT), an LRS platform. Guided by initial conventional PCR screening, we performed ONT sequencing on 34 Adenovirus positive DNA samples, without prior selection based on faeces freshness level or DNA quality. Non-parametric correlation analysis showed that ONT sequencing outputs is not significantly associated (p > 0.05) with DNA concentrations, faeces freshness levels and the OD ratios of A260/A280 and A260/A230. This indicated that conventional DNA quality parameters may not be the most critical factors in determining the suitability of samples for ONT sequencing. A total of 61.76 % (21/34) of the positive-by-PCR-screening samples yielded Adenovirus reads while 38.24 % (13/34) did not in the PCR-free ONT workflow, although rarefaction analysis showed that sequencing saturation was achieved by all samples. Among the 21 samples with adenovirus reads, ten resulted in at least one Adenovirus contig by the Flye assembler while nine did not and two samples had only a single Adenovirus read. Identity similarity above 90 % in conventional PCR screening may help in selecting ONT positive samples.
Consensus on the optimal high-throughput sequencing (HTS) approach to examine biodiversity in mixed terrestrial arthropod samples has not been reached. Metatranscriptomics could increase the proportion of taxonomically informative mitochondrial reads in HTS outputs but has not been investigated for terrestrial arthropod samples. We compared the efficiency of 16S rRNA metabarcoding, metagenomics and metatranscriptomics for detecting species in a mixed terrestrial arthropod sample (pooled DNA/RNA from 38 taxa). 16S rRNA metabarcoding and nuclear rRNA-depleted metatranscriptomics had the highest detection rate with 97% of input species detected. Based on cytochrome c oxidase I, metagenomics had the highest detection rate with 82% of input species detected, but metatranscriptomics produced a larger proportion of reads matching (Sanger) reference sequences. Metatranscriptomics with nuclear rRNA depletion may offer advantages over metabarcoding through reducing the number of spurious operational taxonomic units while retaining high detection rates, and offers natural enrichment of mitochondrial sequences which may enable increased species detection rates compared with metagenomics.