Comments from curator
Phenotyping data comes from the
Supporting Information of the paper (
journal page, also on
PMC; files taken from the PMC Open Access copy) and from
Table 1 of the paper:
-
Table S1 (
supp_g3.115.022558_TableS1.pdf): median time to death of 116 RAL lines injected with 1000 CFU of
L. monocytogenes.
-
Table S2 (
supp_g3.115.022558_TableS2.pdf): bacterial load 48 h after the same injection, 112 lines.
-
Table 1 and
Table S4 (
supp_g3.115.022558_TableS4.pdf): logistic growth parameters of
L. monocytogenes after injection of 100 CFU, and median time to death, for 11 lines.
Only these per-line values were published (no per-fly data), so the study has a summary dataset only. All flies were males (5-7 days old, kept in vials of 20 on dextrose-cornmeal-yeast medium at 25 °C). The RAL lines are uploaded as DGRP lines (RAL_21 = DGRP_021). There was no GWAS.
Screen (Tables S1 and S2): flies were injected in the abdomen with 50 nL containing about 1000 CFU of
L. monocytogenes (strain 10403S).
-
md_MTD_Lm1000: median time to death (days), from daily mortality checks (60 flies per line according to the Methods, 70 according to the Results). It ranges from 1 day (DGRP_321) to 9 days, as in the paper.
-
md_LogCFU48h_Lm1000: log10 of the median number of bacteria per fly (CFU) 48 h after the injection, from single flies homogenised and plated (at least 8 flies). The caption of Table S2 ("Median percent of life shortened") is wrong: its values are log10 CFU per fly, as shown by Table S3, which gives the CFU of 11 of these lines (equal to 10
value, rounded), and by the range given in the text (6.44 × 10
3 for DGRP_359 to 2.41 × 10
6 for DGRP_059).
Growth curves (Table 1 and Table S4): 11 lines with contrasting phenotypes in the screen were injected with about 100 CFU, and the bacterial load per fly was measured at 7 time points during the first 2 days, then daily until the median time to death. The authors fitted a logistic curve to the natural log of the loads (Prism 6; Y = YM·Y0/((YM−Y0)·exp(−k·t) + Y0)), using the data up to the median time to death of the line:
-
mn_GrowthK_Lm100 and
se_GrowthK_Lm100: growth plateau K (YM; ln CFU per fly) and its standard error;
-
mn_GrowthR_Lm100 and
se_GrowthR_Lm100: maximal growth rate r (k; ln CFU per hour) and its standard error;
-
md_MTD_Lm100: median time to death (hours) after the injection of 100 CFU (Table 1).
K and r are model estimates for the line (one fit on all the flies of the line), uploaded as means. The values are taken at full precision from Table S4; Table 1 gives the same values rounded. The fit was poor for the three lines with the lowest plateau (DGRP_359, DGRP_787 and DGRP_821; R
2 0.20-0.35, against 0.72-0.91 for the others), so their growth rates are imprecise; the authors excluded them from some correlations. With these values, the correlations of Figure 4 are reproduced exactly (median time to death vs K: P = 0.0067 with the 11 lines, P = 0.04 without the 3 low-plateau lines; vs r: P = 0.96, R
2 = 3 × 10
-4, and P = 0.003, R
2 = 0.793).
Of note:- The text says that 114 lines were screened, but Table S1 has 116 lines and Table S2 has 112 (123 lines in all: 11 lines are only in Table S1 and 7 only in Table S2). All are uploaded.
- RAL_712 is listed twice in Table S2, with two different values (5.78 and 5.36), so one of them belongs to another line. Its bacterial load is set to NA.
- In Tables S4 and S5, the column labelled "RAL 774" is RAL 508: its values are those of RAL 508 in Table 1, and RAL 508 (not RAL 774) is one of the 11 lines in the Methods, Table S3 and Figure 1. It is uploaded as DGRP_508.
Not uploaded: the growth curves of
w1118 and of the
kenny and
CG2247 piggyBac mutants (Table 3; not DGRP lines); the fits of the same growth curves cut at 72 h (Table S5); the dose-response of RAL 309, RAL 508 and
w1118 injected with 10 to 10
5 CFU (Table 2 and Table S6; only 2 DGRP lines); the goodness of fit and fitted initial loads (Y0) of the logistic curves; and Table S3, which repeats Tables S1 and S2 for the 11 lines.
