Our love-hate relationship with DNA barcodes, the Y2K problem, and the search for next generation barcodes.

Jeffrey M Marcus

doi:10.3934/genet.2018.1.1

Abstract

DNA barcodes are very useful for species identification especially when identification by traditional morphological characters is difficult. However, the short mitochondrial and chloroplast barcodes currently in use often fail to distinguish between closely related species, are prone to lateral transfer, and provide inadequate phylogenetic resolution, particularly at deeper nodes. The deficiencies of short barcode identifiers are similar to the deficiencies of the short year identifiers that caused the Y2K problem in computer science. The resolution of the Y2K problem was to increase the size of the year identifiers. The performance of conventional mitochondrial COI barcodes for phylogenetics was compared with the performance of complete mitochondrial genomes and nuclear ribosomal RNA repeats obtained by genome skimming for a set of caddisfly taxa (Insect Order Trichoptera). The analysis focused on Trichoptera Family Hydropsychidae, the net-spinning caddisflies, which demonstrates many of the frustrating limitations of current barcodes. To conduct phylogenetic comparisons, complete mitochondrial genomes (15 kb each) and nuclear ribosomal repeats (9 kb each) from six caddisfly species were sequenced, assembled, and are reported for the first time. These sequences were analyzed in comparison with eight previously published trichopteran mitochondrial genomes and two triochopteran rRNA repeats, plus outgroup sequences from sister clade Lepidoptera (butterflies and moths). COI trees were not well-resolved, had low bootstrap support, and differed in topology from prior phylogenetic analyses of the Trichoptera. Phylogenetic trees based on mitochondrial genomes or rRNA repeats were well-resolved with high bootstrap support and were largely congruent with each other. Because they are easily sequenced by genome skimming, provide robust phylogenetic resolution at various phylogenetic depths, can better distinguish between closely related species, and (in the case of mitochondrial genomes), are backwards compatible with existing mitochondrial barcodes, it is proposed that mitochondrial genomes and rRNA repeats be used as next generation DNA barcodes.

Highlights

The analysis focused on Trichoptera Family Hydropsychidae, the net-spinning caddisflies, which demonstrates many of the frustrating limitations of current barcodes
Because they are sequenced by genome skimming, provide robust phylogenetic resolution at various phylogenetic depths, can better distinguish between closely related species, and, are backwards compatible with existing mitochondrial barcodes, it is proposed that mitochondrial genomes and ribosomal RNA (rRNA) repeats be used as generation deoxyribonucleic acid (DNA) barcodes
While several other authors have suggested that generation sequencing will have profound effects on how barcoding is conducted [13,26,53,96], this is the first explicit proposal that genome skimming by generation shotgun sequencing be used to enlarge conventional DNA barcode identifiers and upgrade current barcoding strategies

Summary

Introduction

A piece of the mitochondrial cytochrome c oxidase subunit I (COI) gene (Table 1) was employed as the barcode sequence because of the availability of degenerate PCR primers that had been shown to consistently amplify a homologous DNA fragment in diverse organisms [2]. Variable portions of the 16S rRNA are used for barcoding [4,5] In plants, regions of 2 different chloroplast genes are used as a combinatorial barcode: the large subunit of ribulose bisphosphate carboxylase (rbcL) and maturaseK (matK) [6]. The choice and size of each of these barcode regions was based on the technologies for PCR amplification and DNA sequencing that were widespread when the barcode primers were designed [7]. The suggested approach to overcoming these challenges should be broadly applicable to many groups of organisms

Methods

Results

Discussion

Conclusion