Dataset: dataset1_Krug_et_al_2015: Suppl. Table 3: Accessions

For full details on sampling and analytical methodology, see Krug et al, Systematic Biology, 2015.

The data were for a global study of diversity in clade Sacoglossa; all temperate and tropical oceans were sampled. Specimens came from the authors' collections or from museum collections.

Excerpt: "For one to two exemplars per taxon, portions of four loci were sequenced: (i) COI; (ii) mitochondrial large ribosomal subunit rRNA (16S); (iii) nuclear histone III (H3); and (iv) nuclear large ribosomal subunit rRNA (28S). Amplifications and sequencing followed published protocols, with 28S amplified as three overlapping fragments and assembled prior to alignment (Krug et al. 2008; Händeler et al. 2009). Data matrix completeness was 94% (840 cells, locus × taxon). Initial alignments of all loci were done using MUSCLE with default settings in Geneious v6.1.6. Based on published models for rRNA genes (Lydeard et al. 2000; Medina and Walsh 2000; Mallatt et al. 2010), we developed secondary structure models to refine alignments of 16S (Supplementary Figs. S1 and S2) and 28S (Supplementary Fig. S3). Adjustments were made by eye to maintain predicted base pairing interactions in stem regions conserved across Mollusca. Loop regions of ambiguous alignment were removed, as were sequence blocks masked by the least stringent criteria in Gblocks v.0.91b (Castresana 2000). Final aligned sequence partitions were 658 bp (COI), 404 bp (16S), 1392 bp (28S), and 328 bp (H3); NCBI accession numbers are given in Supplementary Table S3. Individual gene trees were built for all loci using Bayesian Inference (BI) and maximum likelihood (ML) as detailed below, to ensure no rogue sequences or unstable taxa were included; topologies were consistent among gene trees except in unresolved regions (Supplementary Fig. S4a-d)."