Table 1.
Description of the ancient samples analysed in this study.
Fig 1.
Overall experimental workflow.
Experimental conditions were investigated in six main tests: Pre-extraction CT-scanning ‘Radiation’, DNA ‘Extraction’, pretreatments such as ‘Digestion’ and ‘Bleach’ wash, as well as the use of the shell ‘Periostracum’, and the aragonite or calcite ‘Shell layer’ as substrate. Ancient mussel shell DNA recovery was compared in these varying conditions by processing five shells per test, for which a fragment was sub-sampled and powdered before being evenly split into aliquots and subjected to DNA extraction, DNA library construction, PCR amplification, purification and HTS.
Fig 2.
Ancient shell DNA sequencing reads were mapped to the North-European Mytilus edulis nuclear reference genome (MeduEUN) [68]. For a given dataset, parameter estimates calculated from full sequencing datasets are represented by coloured points, error bars correspond to the minimum and maximum parameter estimates calculated from ten down-sampled datasets. Light-grey dashed lines link paired samples (i.e., shell powder aliquots from same sample undergoing different treatments). Statistically significant results are indicated with one (p-value < 0.05) or two (p-value < 0.01) asterisk(s). (A) Endogenous DNA content. (B) Sequence clonality estimated from ten down-samples. (C) GC-content.
Fig 3.
Ancient shell DNA sequencing reads were mapped to the North-European Mytilus edulis nuclear reference genome (MeduEUN) [68]. For a given dataset, parameter estimates calculated from full sequencing datasets are represented by coloured points, error bars correspond to the minimum and maximum parameter estimates calculated from ten down-sampled datasets. Light-grey dashed lines link paired samples (i.e., shell powder aliquots from same sample undergoing different treatments). Statistically significant results are indicated with one (p-value < 0.05) or two (p-value < 0.01) asterisk(s). (A) Average fragment length. (B) C-to-T misincorporation rates at the terminal base position of 5’-ends.
Fig 4.
Ancient shell DNA quantity and quality estimates according to specimen age and presence of the periostracum.
Each point represents the mean estimate value averaged across conditions for a given dated specimen, with (N = 14) or without a preserved periostracum (N = 8). (A) Endogenous DNA content. (B) C-to-T misincorporation rates at the terminal base position of 5’-ends. (C) Average fragment length.
Fig 5.
Mitochondrial-to-nuclear DNA ratio corrected by genome sizes.
Ancient shell DNA sequencing reads were aligned to the Mytilus edulis nuclear (MeduEUN) [68] and mitochondrial reference genomes (178mc10) [73] through competitive mapping. For a given dataset, parameter estimates calculated from full sequencing datasets are represented by coloured points, error bars correspond to the minimum and maximum parameter estimates calculated from ten down-sampled datasets. Light-grey dashed lines link paired samples (i.e., shell powder aliquots from same sample undergoing different treatments). Statistically significant results are indicated with one (p-value < 0.05) or two (p-value < 0.01) asterisk(s).