Evaluating intraspecific genetic diversity of a fish population using environmental DNA: An approach to distinguish true haplotypes from erroneous sequences

Abstract

Recent advances in environmental DNA (eDNA) analysis using high-throughput sequencing (HTS) provide a non-invasive way to evaluate the intraspecific genetic diversity of aquatic macroorganisms. However, erroneous sequences present in HTS data can result in false positive haplotypes; therefore, reliable strategies are necessary to eliminate such erroneous sequences when evaluating intraspecific genetic diversity using eDNA metabarcoding. In this study, we propose an approach combining denoising using amplicon sequence variant (ASV) method and the removal of haplotypes with low detection rates. A mixture of rearing water of Ayu (Plecoglossus altivelis altivelis) was used as an eDNA sample. In total, nine haplotypes of Ayu mitochondrial D-loop region were contained in the sample and amplified by two-step tailed PCR. The 15 PCR replicates indexed different tags were prepared from the eDNA sample to compare the detection rates between true haplotypes and false positive haplotypes. All PCR replications were sequenced by HTS, and the total number of detected true haplotypes and false positive haplotypes were compared with and without denoising using the two types of ASV methods, Divisive Amplicon Denoising Algorithm 2 (DADA2) and UNOISE3. The use of both ASV methods considerably reduced the number of false positive haplotypes. Moreover, all true haplotypes were detected in all 15 PCR, whereas false positive haplotypes had detection rates varying from 1/15 to 15/15. Thus, by removing haplotypes with lower detection rates than 15/15, the number of false positive haplotypes were further reduced. The approach proposed in this study successfully eliminated most of false positive haplotypes in the HTS data obtained from eDNA samples, which allowed us to improve the detection accuracy for evaluating intraspecific genetic diversity using eDNA analysis.

Increasing confidence for discerning species and population compositions from metabarcoding assays of environmental samples: case studies of fishes in the Laurentian Great Lakes and Wabash River

“…Aquarium eDNA experiments by Tsuji et al (2018) detected mtDNA control region haplotypes of the ayu Plecoglossus altivelis, but despite DADA2's denoising algorithm (albeit without any frequency based bioinformatic filtering), they discovered 31 false haplotypes, with seven occurring across all 15 replicates. This likely was due to error being non-random on the HTS platform, posing challenges for gathering population genetics data (Nakamura et al 2011;Schirmer et al 2016).…”

“…Better protocols are needed to alleviate primer bias and error in assay data, while correctly identifying as many taxa as possible (Zinger et al 2019). Inaccurate base calls in HTS can artificially inflate population diversity estimated from intra-specific haplotypic diversity (Tsuji et al 2018), which may pose difficulty in distinguishing signal (correct haplotypes) from noise ("false" haplotypes).…”

Population differentiation from environmental DNA: Investigating the potential of haplotype presence/absence‐based analysis of molecular variance

“…Second, our simulations assume that all observed haplotype are real, none being due to PCR artifacts or sequencing errors. Recent improvements have been made toward distinguishing real from spurious haplotypes due to PCR or sequencing errors (Tsuji et al., 2019; Turon et al., 2020). How these spurious haplotypes could impact our approach was not investigated and further research is needed to assess the resulting risk of type I errors and true negatives.…”

“…(c) Which mutations are PCR or sequencing errors and which mutations represent true genetic variability? Recent research has made progress toward answering the second and third questions (Hajibabaei et al., 2011; Tsuji et al., 2019), but there remains no current solution to the first one, and there possibly never will be, unless cells can be isolated from the environment and sequenced individually (Czechowska et al., 2008). Turon et al.…”

“…(c) Which mutations are PCR or sequencing errors and which mutations represent true genetic variability? Recent research has made progress toward answering the second and third questions (Hajibabaei et al, 2011;Tsuji et al, 2019), but there remains no current solution to the first one, and there possibly never will be, unless cells can be isolated from the environment and sequenced individually (Czechowska et al, 2008). Turon et al (2020) tried to overcome this challenge by using a semi-quantitative approach considering the ranking of read frequencies rather than the read frequencies themselves in order to perform an analysis of molecular variance (AMOVA).…”