Testing Efficacy of Assembly-Free and Alignment-Free Methods for Species Identification Using Genome Skims, with Patellogastropoda as a Test Case

Most recently, species identification has leaped from DNA barcoding into shotgun sequencing-based genome skimming alternatives. Genome skims have mainly been used to assemble organelle genomes, which discards much of the nuclear genome. Recently, an alternative approach was proposed for sample identification, using unassembled genome skims, which can effectively improve phylogenetic signal and identification resolution. Studies have shown that the software Skmer and APPLES work well at estimating genomic distance and performing phylogenetic placement in birds and insects using low-coverage genome skims. In this study, we use Skmer and APPLES based on genome skims of 11 patellogastropods to perform assembly-free and alignment-free species identification and phylogenetic placement. Whether or not data corresponding to query species are present in the reference database, Skmer selects the best matching or closest species with COI barcodes under different sizes of genome skims except lacking species belonging to the same family as a query. APPLES cannot place patellogastropods in the correct phylogenetic position when the reference database is sparse. Our study represents the first attempt at assembly-free and alignment-free species identification of marine mollusks using genome skims, demonstrating its feasibility for patellogastropod species identification and flanking the necessity of establishing a database to share genome skims.

Automated summary

Recently, species identification has advanced from DNA barcoding to genome skimming based on shotgun sequencing. Genome skimming has been primarily used for assembling organelle genomes, while discarding most of the nuclear genome. A new approach using unassembled genome skims has been proposed for sample identification, which effectively improves phylogenetic signal and identification resolution. Skmer and APPLES have been shown to work well in estimating genomic distance and performing phylogenetic placement using low-coverage genome skims of birds and insects. In this study, Skmer and APPLES were used to perform assembly-free and alignment-free species identification and phylogenetic placement of 11 patellogastropods. Skmer selects the best matching or closest species with COI barcodes, except for species belonging to the same family as the query, under different sizes of genome skims, regardless of whether the data corresponding to the query species are present in the reference database. When the reference database is sparse, APPLES cannot accurately place patellogastropods in the correct phylogenetic position. This study demonstrates the feasibility of genome skims for patellogastropod species identification and highlights the need for establishing a database to share genome skims.