Influence of ploidy, sex and metabolic activity on the quality of de novo genome assemblies – a test using eastern carpenter bees

In haplodiploid insects, males are haploid and females are diploid, providing opportunities for evaluating how the higher heterozygosity of diploids influences the quality of genome assemblies. We did this by comparing genome assembly quality for five males and two females of the eastern carpenter bee, Xylocopa virginica. We used wing wear scores as an index of flight metabolic activity and potential for oxidative damage to flight muscles. DNA was extracted from thoracic flight muscles of adult bees with no visible parasites. We used Illumina sequencing and SoapDenovo2 based on standardized kmer sizes to assemble genomes, and used Quast to calculate assembly quality metrics. We compared assembly quality metrics of males vs. females using a relative quality ratio, R = (average male metric / average female metric). Male genome assemblies were dramatically better than female assemblies: they comprised far fewer contigs (R=0.08), had many fewer mismatches (R=0.04), much higher N50 values (R=55), and much lower L50 values (R=0.04). Male assemblies also tended to require less computing time. Comparisons of two worn and two unworn individuals showed their genome assemblies were of similar quality. In a preliminary analysis we found that the density of variants was moderately correlated with the density of contig/scaffold breakpoints. This suggested that somatic DNA damage may have a small negative effect on assembly quality. Our results support the prediction that genetic heterozygosity resulting from diploidy and possible from somatic DNA variability can negatively affect the quality of an assembly, with ploidy playing the larger role.