Fig 1.
Workflow for recovering DNA from water samples.
Bathwater or other water samples were filtered through glass microfiber filters, cut into pieces, digested with Proteinase K, and subjected to DNA extraction, followed by STR analysis. Filtration was repeated until the entire sample volume was processed (up to eight successive filtration cycles).
Table 1.
Cases in which human DNA was extracted from water samples collected at drowning sites.
Fig 2.
Box–violin plots showing (A) the quantity of human DNA in bathwater and (B) the DNA degradation index at each immersion time.
Dotted lines indicate median values. Variables in (A) and (B) are shown on a logarithmic scale. Notches represent the 95% confidence interval for the median. A higher degradation index reflects greater DNA degradation. Statistical significance was assessed by one-way ANOVA followed by the Steel–Dwass test (n = 11, *p < 0.05, **p < 0.01).
Fig 3.
Box–violin plots showing the percentage of loci classified as (A) matching reference profile, (B) allelic non-detection, (C) allelic mixtures, (D) allele drop-ins, and (E) only non-bather loci at each immersion time.
Dotted lines indicate median values. Notches represent the 95% confidence interval for the median. Statistical significance was assessed by one-way ANOVA followed by the Steel–Dwass test (n = 11, *p < 0.05, **p < 0.01).
Fig 4.
Changes in (A) the quantity of human DNA in bathwater, and (B) the percentage of loci matching the bather’s reference profile before and after immersion for 1, 2, 5, and 10 minutes in each volunteer.
Panel (A) is plotted on a logarithmic scale. Symbols represent individual values, error bars indicate the 95% confidence interval for the median, and dotted lines connect the median values across immersion times (n = 11).
Table 2.
Quantities of human DNA, DNA degradation indices, and STR profiling results from water samples collected at drowning sites. Case numbers correspond to those in Table 1.