Counting with DNA in metabarcoding studies: How should we convert sequence reads to dietary data?

Deagle BE, Thomas AC, McInnes JC, Clarke LJ, Vesterinen EJ, Clare EL, Kartzinel TR, Eveson JP. · 2019

Read Counting with DNA in metabarcoding studies: How should we convert sequence reads to dietary data? by Deagle BE, Thomas AC, McInnes JC, Clarke LJ, Vesterinen EJ, Clare EL, Kartzinel TR, Eveson JP. (2019) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc6905394.txt.

Introduction

Deagle and co-authors examine a measurement problem in Molecular Ecology. DNA metabarcoding recovers food DNA from droppings or gut contents, revealing what an animal ate. Samples yield counts of sequences per food type. What those counts mean is disputed. Many researchers trust only presence-or-absence data, calling it the safer, more conservative choice.

Sequence recovery is biased: some foods amplify better than others, so counts can mislead. The authors explore real dietary datasets to see how badly, and when relative counts still track the true diet. Getting the true diet by any method is hard, they stress. Their guidance weighs correction factors against study goals, rather than banning counts outright.

Catalog record

CREATORDeagle BE, Thomas AC, McInnes JC, Clarke LJ, Vesterinen EJ, Clare EL, Kartzinel TR, Eveson JP.
YEAR2019
COLLECTIONPAPERS
FORMATTXT
FILE SIZE74 KB
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC6905394

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Counting with DNA in metabarcoding studies: How should we convert sequence reads to dietary data? By Bruce E. Deagle, Austen C. Thomas, Julie C. McInnes, Laurence J. Clarke, Eero J. Vesterinen, Elizabeth L. Clare, Tyler R. Kartzinel, J. Paige Eveson Published in: Molecular Ecology DOI: 10.1111/mec.14734 ABSTRACT Abstract Advances in DNA sequencing technology have revolutionized the field of molecular analysis of trophic interactions, and it is now possible to recover counts of food DNA sequences from a wide range of dietary samples. But what do these counts mean? To obtain an accurate estimate of a consumer's diet should we work strictly with data sets summarizing frequency of occurrence of different food taxa, or is it possible to use relative number of sequences? Both approaches are applied to obtain semi‐quantitative diet summaries, but occurrence data are often promoted as a more conservative and reliable option due to taxa‐specific biases in recovery of sequences. We explore representative dietary metabarcoding data sets and point out that diet summaries based on occurrence data often overestimate the importance of food consumed in small quantities (potentially including low‐level contaminants) and are sensitive to the count threshold used to define an occurrence. Our simulations indicate that using relative read abundance (RRA) information often provides a more accurate view of population‐level diet even with moderate recovery biases incorporated; however, RRA summaries are sensitive to recovery biases impacting common diet taxa. Both approaches are more accurate when the mean number of food taxa in samples is small. The ideas presented here highlight the need to consider all sources of bias and to justify the methods used to interpret count data in dietary metabarcoding studies. We encourage researchers to continue addressing methodological challenges and acknowledge unanswered questions to help spur future investigations in this rapidly developing area of research. INTRODUCTION Many recent studies documenting trophic interactions make use of metabarcoding, an approach which combines high‐throughput sequencing (HTS) with DNA barcoding to identify the food remains present in faecal samples or stomach contents (Nielsen, Clare, Hayden, Brett, & Kratina, 2017). When HTS first became available, the potential applications in diet studies were clear and the methods were quickly embraced by the community (Deagle, Kirkwood, & Jarman, 2009; Valentini et al., 2009). In a comprehensive review of DNA‐based diet analysis by King, Read, Traugott, and Symondson (2008), the possibility of using HTS was only briefly mentioned as a “Future Direction,” and just four years later, another review paper focussed entirely on this approach (Pompanon et al., 2012). While many underlying technical and biological details vary between dietary metabarcoding studies, the general workflow is now well defined. It involves extraction of total DNA from the dietary sample, PCR amplification of DNA barcode markers from food taxa of interest and then DNA sequencing for taxonomic classification of the recovered sequences. The workflow has been applied to determine diet in a range of animals, from invertebrates to large mammalian herbivores and carnivores (representative studies summarized in Table 1). [Table: Use of sequence counts in 20 metabarcoding diet studies carried out using faecal DNA collected from a range of different species. Representative studies across a range of focal taxa carried out by different research groups are shown rather than trying to summarize all dietary metabarcoding studies] The rapid adoption of HTS to characterize complex mixtures of DNA is not unique to dietary studies; over the last ten years, the technology has produced a wealth of new genetic data providing insight into almost all areas of biology (Goodwin, McPherson, & McCombie, 2016). One feature of HTS is that it provides counts of DNA sequences in each sample, and therefore, it has the potential not only to provide a qualitative list, but also to quantify what DNA is present. The interpretation of sequence read counts as a numerical representation of sample composition is common in many HTS applications. For example, studies sequencing transcripts to determine differences in gene expression (Finotello & Di Camillo, 2015), profiling microbe communities (Vandeputte et al., 2017) or measuring epigenetic variation (Schield et al., 2016) all rely on sequence read counts. However, decisions about how to interpret read counts is certainly not routine and the validity of interpretations is sometimes questioned even in fields where the practice is well established (e.g., Edgar, 2017; Olova et al., 2017). These debates are constructive and should motivate researchers to test underlying assumptions and justify their interpretations, but can give rise to the impression that count data are always misleading. The reality is that all metabarcoding studies use sequence counts to some extent. In dietary investigations, count data are used either to record the occurrence of food species within samples based on a threshold number of sequences (i.e., presence/absence of taxa), or to calculate the percentage of DNA belonging to each food species as a proxy for relative biomass consumed (i.e., relative abundance of taxa; Figure 1). The conversion of sequence counts to occurrence data is often considered a more conservative approach than using proportional data. In their introduction to the Molecular Ecology Special Issue on “Molecular Detection of Trophic Interactions”, Symondson and Harwood (2014) pointed out that authors of many metabarcoding papers “now simply record numbers of predators testing positive for a target prey or plant species, providing a pragmatic and useful surrogate for truly quantitative information.” This sentiment, that focusing only on occurrence data is a reliable and safe option, is now common in the literature, and this step in the analysis pipeline is often uncritically applied as the default option. Using counts as an indication of biomass in samples is more controversial. Indeed, the difficulties of obtaining an accurate biomass signature from sequence counts include both technical and biological biases that affect barcode marker recovery rates from different taxa (Amend, Seifert, & Bruns, 2010; Deagle et al., 2009; Pompanon et al., 2012). Therefore in the best‐case scenario sequence read counts can only provide a rough estimate of proportional abundance. Still, to accept the notion that relative sequence counts provide no meaningful information would mean that, within one sample, a few DNA sequences from one food taxon are equivalent to 10,000 sequences from another. Most molecular ecologists would interpret these disparate counts to mean that there are differences in template DNA abundance (as long as methods used to collect the data are reasonable) and that there is some biological basis for that difference. Ignoring this difference may inhibit ecological understanding. [Figure: Information in faecal samples from dietary metabarcoding data sets of an albatross (McInnes, Jarman, et al., 2017), an insectivorous bat (Vesterinen et al., 2016) and Grevy's zebra (Kartzinel et al., 2015). (a) Individual‐level data in 10 faecal samples viewed using different metrics. Colours represent different food taxa. (b) Population‐level summaries of these data sets showing the top 15 food taxa (%FOO ranking); 1% threshold used for occurrence in POO and wPOO calculations. In the lower plots, sum contribution of remaining food taxa is plotted as final point. In each example, population data include only collections from one site and samples with >50 food taxa reads; the albatross data only consider the fish component of the diet (i.e., fish‐specific PCR primers) [Colour figure can be viewed at http://wileyonlinelibrary.com]] Here, we review the approaches taken to interpret

Read the complete text