Are we there yet? Benchmarking low-coverage nanopore long-read sequencing for the assembling of mitochondrial genomes using the vulnerable silky shark Carcharhinus falciformis

Abstract

Background: Whole mitochondrial genomes are quickly becoming markers of choice for the exploration of within-species genealogical and among-species phylogenetic relationships. Most often, 'primer walking' or 'long PCR' strategies plus Sanger sequencing or low-pass whole genome sequencing using Illumina short reads are used for the assembling of mitochondrial chromosomes. In this study, we first confirmed that mitochondrial genomes can be sequenced from long reads using nanopore sequencing data exclusively. Next, we examined the accuracy of the long-reads assembled mitochondrial chromosomes when comparing them to a 'gold' standard reference mitochondrial chromosome assembled using Illumina short-reads sequencing.

Results: Using a specialized bioinformatics tool, we first produced a short-reads mitochondrial genome assembly for the silky shark C. falciformis with an average base coverage of 9.8x. The complete mitochondrial genome of C. falciformis was 16,705 bp in length and 934 bp shorter than a previously assembled genome (17,639 bp in length) that used bioinformatics tools not specialized for the assembly of mitochondrial chromosomes. Next, low-pass whole genome sequencing using a MinION ONT pocket-sized platform plus customized de-novo and reference-based workflows assembled and circularized a highly accurate mitochondrial genome in the silky shark Carcharhinus falciformis. Indels at the flanks of homopolymer regions explained most of the dissimilarities observed between the 'gold' standard reference mitochondrial genome (assembled using Illumina short reads) and each of the long-reads mitochondrial genome assemblies. Although not completely accurate, mitophylogenomics and barcoding analyses (using entire mitogenomes and the D-Loop/Control Region, respectively) suggest that long-reads assembled mitochondrial genomes are reliable for identifying a sequenced individual, such as C. falciformis, and separating the same individual from others belonging to closely related congeneric species.

Conclusions: This study confirms that mitochondrial genomes can be sequenced from long-reads nanopore sequencing data exclusively. With further development, nanopore technology can be used to quickly test in situ mislabeling in the shark fin fishing industry and thus, improve surveillance protocols, law enforcement, and the regulation of this fishery. This study will also assist with the transferring of high-throughput sequencing technology to middle- and low-income countries so that international scientists can explore population genomics in sharks using inclusive research strategies. Lastly, we recommend assembling mitochondrial genomes using specialized assemblers instead of other assemblers developed for bacterial and/or nuclear genomes.

Keywords: Elasmobranch; Long-read sequencing; Nanopore.

Fig. 1

Circularized mitochondrial genome ideogram of the silky shark Carcharhinus falciformis. The map is annotated and depicts a single putative control region, 22 transfer RNA (tRNA) genes, 2 ribosomal RNA genes (rrnS [12S ribosomal RNA] and rrnL [16S ribosomal RNA]), and 13 protein-coding genes (PCGs). Shark photograph: Joi Ito (Attribution 2.0 Generic [CC BY 2.0])

Fig. 2

Sequence errors per de novo (Unicycler and Flye) and reference-based assemblers (Rebaler) without and with ‘extra polishing’ using the program Medaka for the silky shark Carcharhinus falciformis mitochondrial genome. Benchmarking of all long-read assemblies occurred against the Illumina short-read assembly (‘gold’ standard)

Fig. 3

Annotation of reference-based (Rebaler) and de novo (Fyer and Unicycler) mitochondrial genomes assembled using long reads in the silky shark Carcharhinus falciformis. Assemblies depicted include those with and without ‘extra polishing’ with the program Medaka

Fig. 4

Mitophylogenomic analysis of the genus Carcharhinus and allies, including mitochondrial genomes of the silky shark Carcharhinus falciformis assembled with long reads exclusively and short reads (‘gold standard’). Nodes with bootstrap support values > 90 are marked with an orange circle. Shark photograph: Joi Ito (Attribution 2.0 Generic [CC BY 2.0])

Fig. 5

Barcoding analysis of the genus Carcharhinus using the D-Loop/Control Region (CR), including the CR retrieved from mitochondrial genomes of the silky shark Carcharhinus falciformis mitochondrial genome assembled with long reads alone and short reads (‘gold standard’) plus 447 other specimens belonging to the genus Carcharhinus retrieved from Genbank. Shark drawings from M. Dando (used with permission) [47]

Fig. 6

Bioinformatics pipeline to assemble the mitochondrial chromosome of the silky shark Carcharhinus falciformis using nanopore long reads exclusively