PCR and Illumina high-throughput sequencing
Based on about 200 DNA sequences from a library of A. marinaexpressed sequence tags (Huang et al., 2014), we developed a new set of primers anchored at exons but spanning at least one intron. The 94 pairs of primers producing amplicons 500 to 1500 bps long were used in this study. We performed polymerase chain reaction (PCR) amplification on DNA pool from each population using our 94 primer pairs. To reduce amplification errors, TaKaRa high-fidelity PrimerStar HS DNA polymerase was used. The 30 μL PCR mixture consists of 3 μL 10x TaqBuffer (Mg2+), 3 μL dNTPs (2mM/μL), 1.5 μL of each primer (10μM/μL), 0.5 μL HS DNA Polymerase, 3 μL DNA template (~10ng/μL) and 19 μL deionized water. The PCR program was: 4 min at 94°C; 30 cycles of 10 s at 94°C, 30 s of annealing at the corresponding temperature (Table S1 in the online supplementary file), extension at 72°C for 2 min; followed by 8 min final extension at 72°C. Reactions were held at 16°C before PCR products were subjected to electrophoresis on 1.2% agarose gels. Target bands were excised under ultraviolet light and extracted using the Pearl DNA Gel Extraction Kit (Pearl, Guangzhou, China). Extracted DNA was examined by NanoDrop 2000 to ensure that the amount of each gene product was no less than 100ng. PCR products of the 94 loci from the same population were again pooled, using 100 ng of DNA per locus. We thus obtained 16 PCR product pools, each including amplicons from 94 loci.
PCR product pools from each population were delivered for sequencing on the Illumina Genome Analyzer and Illumina HiSeq 2000 platform at BGI (Shenzhen) following the manufacturer’s instructions. 200 bp DNA libraries were constructed for these mixtures and an 8 bp index in the adapter was used to distinguish the populations. Method details used for library construction were the same as those detailed in the supplementary materials of our previous publication (Guo et al., 2016). Raw reads produced from the Illumina Genome Analyzer platform were 90 bps in length (all populations except MC, BB, and DW; abbreviations of population names are defined in Table 1), while those from the Illumina HiSeq 2000 platform were 130 bps in length (MC, BB, and DW).