UNBIAS: An attempt to correct abundance bias in 16S sequencing, with limited success

Abstract

Next-generation amplicon sequencing of 16S ribosomal RNA is widely used to survey microbial communities. Alpha and beta diversities of these communities are often quantified on the basis of OTU frequencies in the reads. Read abundances are biased by factors including 16S copy number and PCR primer mismatches which can cause the read abundance distribution to diverge substantially from the species abundance distribution.Using mock community tests with species abundances determined independently by shotgun sequencing, I find that 16S amplicon read frequencies have no meaningful correlation with species frequencies (Pearson coefficient r close to zero). In addition, Ishow that that the Jaccard distance between the abundance distributions for reads of replicate samples, which ideally would be zero, is typically ~0.15 with values up to 0.71 for replicates sequenced in different runs. Using simulated communities, I estimate that the average rank of a dominant species in the reads is 3. I describe UNBIAS, a method that attempts to correct for abundance bias due to gene copy number and primer mismatches. I show that UNBIAS can achieve informative, but still poor, correlations (r ~0.6) between estimated and true abundances in the idealized case of mock samples where species are well known. However, r falls to ~0.4 when the closest reference species have 97% identity and to ~0.2 at 95% identity. This degradation is mostly explained by the increased difficulty in predicting 16S copy number when OTUs have lower similarity with the All rights reserved. No reuse allowed without permission.was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint (which . http://dx.doi.org/10.1101/124149 doi: bioRxiv preprint first posted online Apr. 4, 2017; reference database, as will typically be the case in practice. 16S abundance bias therefore remains an unsolved problem, calling into question the naive use of alpha and beta diversity metrics based on frequency distributions.

Accuracy of microbial community diversity estimated by closed- and open-reference OTUs

“…I constructed a reference database of known tags as follows. I used the SEARCH_16S algorithm (Edgar, 2017a) to search the genome of each strain and identify its full-length 16S rRNA sequence(s), as described in (Edgar, 2017b). For each pair of primers, I extracted the segment between the primer-matching loci.…”

“…To investigate this, I identified the subset ( GG-tagsX ) of GG-tags which has binding loci for all primer pairs. The GG-tagsX subset comprises only 60,470 sequences (4.8% of the full GG database) because most Greengenes sequences are truncated such that they lack binding sequences for the V1 forward primer and V9 reverse primer (Edgar, 2017b). I assigned all tags in GG-tagsX to OTUs by Qclosed and calculated the probability that a given pair of tags in the same sequence would be assigned to the same OTU.…”

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

“…Conclusions from analyses of mock communities vary from no relationship to good correlations between the composition of the mixture and sequence reads (Edgar, ; Kimmerling et al., ; Pornon et al., ). One reason for these different conclusions is that the range of concentrations analysed varies considerably across studies, from nearly equal mixtures of a few taxa, to mixtures containing many taxa in very different abundances.…”

“…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.…”

Consistent and correctable bias in metagenomic sequencing experiments

“…Therefore, we expect extraction efficiencies to vary among taxa, but be approximately constant for any specific taxon across samples treated with the same protocol. Other multiplicative sources of bias include variation in PCR binding and amplification efficiencies (Wagner et al, 1994;Suzuki and Giovannoni, 1996;Polz and Cavanaugh, 1998;Edgar, 2017) and in marker-gene copy number (Kembel et al, 2012).…”