DNA extraction, nSSR genotyping and cpDNA sequencing
DNA extraction . Total genomic DNA was extracted using the MATAB protocol and chloroform separation with slight modifications following Mariac et al . . We checked DNA quality and quantity using a NanoDropTM One spectrophotometer (Thermo ScientificTM). Extracted DNA was diluted in 50 μL Tris-EDTA (TE) buffer (1X) pH 8.0 and stored at −20°C.
nSSR genotyping . All individuals were genotyped using 14 polymorphic nuclear nSSR markers, with PCR conditions as described in Kusuma et al. . As one of the primer pair (Myr 42 locus) allowed for the amplification of two different genomic regions, corresponding information was split into two, namely Myr 42a andMyr 42b (see Kusuma et al ., 2020 for details). From this, we obtained our final dataset of 15 diploid genotypes for each individual. Genotyping was done on an ABI 3500 XL (Applied Biosystems) at CIRAD (Montpellier, France) using 1 μL of PCR product, 12 μL of Hi-Di Formamide (Applied Biosystems) and 0.3 μL GeneScan™ 500 LIZ™ dye Size Standard (Applied Biosystems). Genotypes were analyzed for each locus using the automatic procedure implemented in Geneious Prime® v.2022.2.2 and manually corrected.
cpDNA sequencing . We selected three to five individuals per subpopulation for genomic library construction (N = 90; see Supplementary File 1 for details) as well as four additional individuals of the sister species Myristica argentea (commonly known as the Papuan nutmeg, from Fakfak, Papua, Indonesia) that has been used as an outgroup in phylogenetic network analyses (N = 94 in total; see Supplementary File 1 for details). The protocols for library preparation, chloroplast enrichment, multiplexing, and sequencing are described in Mariac et al . . All sequenced individuals were analyzed as described in Scarcelli et al. . In short, data were demultiplexed and cleaned using the scripts DEMULTADAPT (https://github.com/Maillol/demultadapt) and CUTADAPT v.1.2.1 . Reads were then filtered using a PERL script (Supplementary File 2). After discarding low quality reads, mapping was performed with BWA v.0.7.12 using the M. fragrans cpDNA genome as reference. Unmapped reads were discarded and mapped reads were sorted using SAMTOOLS v.1.9 . Then VARSCAN v.2.3.7 was used to call Single Nucleotide Polymorphisms (SNPs). All SNPs were filtered using VCFtools , by deleting all the indels, and all non-biallelic SNPs. Then, we calculated the mean depth per individual, mean depth per SNP, percentage of missing data per individual, and percentage of missing data per SNP using VCFtools . We provide the command lines and scripts to prepare sequence data, mapping and calling of SNPs in Supplementary File 3.
Genetic diversity and population structure of M. fragrans