DNA extraction

DNA was extracted following two different protocols. In both cases, the dissected tissue was digested overnight in Laird’s buffer and homogenized with 20μl of proteinase K (20mg/ml, >600 mAU/ml), followed by incubation with RNase A at 37°C for 30 minutes. DNA was extracted from thoraces using salt extraction; 300 of NaCl (5M) was added, followed by centrifugation for 15 minutes at 13,000 revolutions per minute (rpm). Three washing steps were completed with one volume of 70% ethanol and centrifuging for five minutes at maximum speed. The remaining pellet was air-dried and then resuspended in 30 μl of MilliQ H₂O. For the abdomens, a phenol-chloroform extraction protocol was used. Two cycles of phenol:chloroform:isoamyl alcohol (25:24:1) addition and centrifugation for five minutes at 13,000 rpm were completed, plus a third cleaning cycle using only chloroform. Precipitation of DNA was achieved by adding 2x volumes isopropanol + 0.1x 3M NaAc, incubating at -18°C overnight and centrifuging for 15 minutes at 13,000 rpm. The final pellet was resuspended with 30 μl of MilliQ H₂O. DNA purity was assessed with NanoDrop, and concentration measured with Qubit DNA Broad Range.

Sequencing

To capture the genetic variation in the population in Sweden, 84 individuals from different geographic regions and the highest DNA quality were selected for analysis. Library preparation for all 84 samples using the TruSeq PCR-free kit followed by multiplexing, and sequencing on two NovaSeq 6000 S4 lanes with 2×150 bp reads, were performed at the National Genomics Infrastructure (NGI), Stockholm.

Read trimming

Illumina sequencing adapters were trimmed by eliminating the first fifteen base pairs (bp) on each end of the raw reads with CutAdapt 1.9.1 (Martin 2011), filtered on Q-score < 30 and a minimum length of 30 bp. Read quality after cleaning was assessed with FastQC (Andrews 2010). Before filtering, an average of 4.3 million reads per sample were obtained, and 2.5% were filtered out.

Mapping and filtering

For each individual, paired-end reads were mapped to the reference genome with bwa v0.7.17 (Li and Durbin 2009). Samtools v1.10 (Li et al. 2009) was used to select reads with paired information. MarkDuplicatesSpark as implemented in GATK v4.1.4.1 (McKenna et al. 2010) was used to eliminate duplicated regions with the –remove-sequencing-duplicates option.

Recombination rate estimation

Recombination rate was estimated with pyrho, and averaged over 1Mb intervals to assess the association of regional recombination rate with several genomic features.