SEARCH

SEARCH BY CITATION

FilenameFormatSizeDescription
men12049-sup-0001-FigS1-S8-TableS1-S6.pdfapplication/PDF173K

Fig. S1 Comparison of the standard estimator that ignores the presence of missing values (i.e. estimator 0T and 0L) against the best performing estimator we proposed (estimator ET and EL).

Fig. S2 Comparison of the MSE for all estimators for low missing value rate scenarios 1–18 and high missing value rate scenarios 19–36 (as described in Table 6 and 7 of the manuscript), for the temporal method (panels a and b) and the LDNe method (panels c and d).

Fig. S3 Comparison of the MSE for estimator D (red line), estimator E (black line) and F (blue line) for low missing value rate scenarios 1–18 and high missing value rate scenarios 19–36 (as described in Table 6 and 7 of the manuscript), for the temporal method (panels a and b) and the LDNe method (panels c and d).

Fig. S4 Comparison of the MSE for estimator E (black line) and F (blue line) for low missing value rate scenarios 1–18 and high missing value rate scenarios 19–36 (as described in Table 6 and 7 of the manuscript), for the temporal method (panels a and b) and the LDNe method (panels c and d).

Fig. S5 Results from the investigation of non-random missing values for estimator D (red line), estimator E (black line) and F (blue line) for low missing value rate scenarios 1–18 and high missing value rate scenarios 19–36 (as described in Table 6 and 7 of the manuscript), for the temporal method (panels a and b) and the LDNe method (panels c and d).

Fig. S6 Results from the investigation of non-random missing values.

Fig. S7 Results from the investigation into using randomly generated population allele frequencies values.

Fig. S8 Results from the investigation into using randomly generated population allele frequencies values.

Table S1 Performance of the temporal method for low missing data rate (0–10%).

Table S2 Performance of the LDNe method for low missing data rate (0–10%).

Table S3 Performance of the temporal method for high missing data rate (0–50%).

Table S4 Performance of the LDNe method for high missing data rate (0–50%).

Table S5 Performance of the temporal method for low missing data rate (0–10%) when samples S1 and S2 were taken at generations 1 and 9.

Table S6 Performance of the temporal method for high missing data rate (0–50%) when samples S1 and S2 were taken at generations 1 and 9.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.