Comments from curator
Phenotyping data come from the authors'
Dryad deposit (
doi:10.5061/dryad.9kj41f0, CC0; downloaded from its
Zenodo mirror, same file and md5), file
Kutzer_et_al_data_files.zip: 13 CSV files named after the statistical models of the paper (models 1-7, described in Appendix S1 of the
paper).
Design: virgin females of 4 DGRP lines (DGRP_350, 367, 379 and 765, chosen from Kutzer et al. 2018,
J Evol Biol 31:159,
DGRPool study 177) received a
primary exposure at about 4 days of age: an injection (18.4 nL in the thorax) of heat-killed
Lactococcus lactis (
L, about 1,840 bacteria), heat-killed
Pseudomonas entomophila (
P, about 1,840 bacteria) or Ringer's solution (
R). Six days later each female was mated once to a male of her own line, and the next day (7 days after the primary exposure) she was
challenged with live
L. lactis (about 1,840 bacteria), live
P. entomophila (about 92 bacteria) or Ringer's solution. The 7 combinations are coded in the phenotype names as <primary><challenge>:
RR,
LR and
PR (Ringer's challenge),
RL,
LL,
RP and
PP (R−R ... P−P in the paper). 25 °C, sugar-yeast-agar medium. Females only (sex F).
Experiment 1 (4 lines, 6 replicates, blocks 1-3 within each replicate):
-
raw_1 (models 2-4; files Exp1_Model_3a/3b (fecundity) and Exp1_Model_2a_4a/2b_4b (resistance_tolerance), joined by fly ID):
FecPre_*, adult offspring produced in the ~26 h between mating and challenge;
FecPost_*, adult offspring produced in the 24 h after the challenge;
CFU1d_*, bacterial load (colony-forming units in a whole-fly homogenate) 1 day post-challenge (DPC), challenged females only. Each row holds one female of each treatment of the same line, replicate and block (different females; row k holds the k-th female by fly ID);
Replicate and
Block are obsolete design columns. 877 females (RR 124, LR 141, PR 140, RL 124, LL 139, RP 107, PP 102); the R−R females are in both fecundity files with identical values and are uploaded once.
-
raw_2 (models 1a, 1b): number of females alive (
Alive1d_*) and dead (
Dead1d_*) 1 DPC, per line, treatment and replicate (
Replicate1d, obsolete); the R−R counts are identical in both files.
-
raw_3 (models 5a-f): expression of
Dscam1,
Drosomycin (
Drs) and
Metchnikowin (
Mtk) 1 DPC, measured by RT-qPCR on single females of replicates 5 and 6 that were frozen instead of being homogenized for the load (treatments RR, LR, RL and LL only). Values are the authors' 2
-ΔCT:
Drs and
Mtk relative to the geometric mean of the CTs of
Rp49 and
RpL13a,
Dscam1 relative to
RpL13a measured with it (as recomputed from the CTs of the files; the
Dscam1 values are given to 3 significant digits). The paper analysed their log. The three genes of a female are on one row: the per-fly correlation file (models 5d-f) links the
Dscam1 file and the
Drs/
Mtk file (182 females: 174 with the three genes, 7 without
Dscam1, 1 with
Dscam1 only). The CTs are not uploaded. Design columns
ReplicateQ and
BlockQ (obsolete).
Experiment 2 (DGRP_379 and DGRP_765 only, 3 replicates of 3 blocks),
raw_4: survival censused daily until 28 DPC (models 6a-c):
SurvTime_*, day of death, or 28 for the females alive at day 28;
Dead_*, 1 = died, 0 = alive at day 28 (censored), from the Censor column of the files. 751 females, 52-54 per line and treatment, stacked as in raw_1. Then, on separate rows (other females), the
L. lactis load of the survivors homogenized at 28 DPC (model 7:
CFU28d_RL,
CFU28d_LL; 29 females, one without bacteria). The loads of the surviving
P. entomophila- and Ringer's-challenged females (no bacteria, according to the paper) are not in the files. Design columns
ReplicateE2 and
BlockE2 (obsolete).
Not uploaded: two derived per-fly columns of the resistance_tolerance files, PercentChange (post-challenge fecundity relative to the mean of the Ringer's-challenged females of the same line and primary exposure) and SlopeG_T (not described); the paper does not use them. Fecundity tolerance is the slope of FecPost on ln(CFU1d) per line and treatment (models 4a, 4b), and resistance the inverse of the load: both can be computed from raw_1. Survival between the primary exposure and mating (88-96% per line) is only given in the text.
Inconsistencies in the files (values kept as published): female 359 (DGRP_379, RL) has no post-challenge fecundity in the fecundity file but 16 offspring in the resistance_tolerance file (16 uploaded). Females 186 (DGRP_367, LL; pre-challenge fecundity NA) and 425 (DGRP_350, RP) are only in the resistance_tolerance files and were added with the fecundity given there. 12 challenged females have fecundity but no load. The number of females alive 1 DPC (raw_2) and the number of females with fecundity data (raw_1) differ by 1-2 in 8 of the 168 line × treatment × replicate groups. Six
P. entomophila-challenged females have a load of 0 CFU, although the paper analysed ln-transformed loads.
Checks: as in the paper, 9/5 (L−L) and 10/5 (R−L) females of RAL379/RAL765 with a 28-DPC load, of mean 4.3 × 10
5 ± 3.3 × 10
5 (SE); 1-DPC
L. lactis load of the same two lines 4.2 × 10
6 ± 1.1 × 10
6; 52-54 females per survival curve; 16-24 females per line and challenge for the gene expression; 958 females in the 1-DPC survival counts (paper: 959 animals).