Comments from curator
Phenotyping data comes from the authors'
GitHub repository (
ben6uw/Harrison_Lee_et-al-2024, given in the Data Availability Statement; the data are on the branches, not on main) and from
Table S1 of the paper (
journal page, also on
PMC):
-
Phenotype_files branch:
fDEVrapa_COMPILED_GENOTYPED.csv (pupae counted per vial and census in the DGRP screen),
fDEV DGRP RAPA genotype summary.csv (the authors' line means) and
food deprivation files/ (
vial_key.csv,
dat.csv: food-dilution experiment).
-
GWAS_files branch:
GWASinput_FINAL.txt (GWAS phenotype).
-
Metabolomics_files branch:
data/lcms.data.csv,
Sample_ID.csv and
Mitchell_sample_meta.data.csv (larval metabolome).
-
Table S1 (ACEL-23-e14292-s003.xlsx): pupation time statistics per line and treatment, and the developmental delay.
Screen design. Embryos of 140 DGRP lines (and of Canton-S and
w1118) were placed on food overlaid with 2 nmol rapamycin in 50 µL ethanol, or with ethanol alone (vehicle control): 40 embryos per vial, three vials per line and treatment, 25 °C. New pupae were counted and removed once or twice a day until no vial produced pupae for two days. Lines were screened in 11 batches of about 10-30 lines, with DGRP_045, DGRP_321, Canton-S and
w1118 in every batch. The larvae were not sexed (sex NA). The authors genotyped the parents by PCR and removed 14 stocks that were not the expected DGRP line; their data are not in the file used here (two of them, 439 and 820, are in raw_2: see below). Canton-S and
w1118 were dropped (not DGRP lines), and so was the Wolbachia column (DGRP annotation).
raw_1 (screen, one row per pupa):
PupTime_Ctrl and
PupTime_Rapa are the pupation times in days, computed as in the authors' scripts: census day + clock time (hhmm)/2400. Day 0 is the day the embryos were collected (in the morning), so the time is counted from midnight. Note that hhmm/2400 is not an exact clock conversion: 10:30 counts as 0.429 d instead of 0.4375 d, so values are up to 0.015 d (about 21 min) early. Pupae of the two treatments are unrelated: row k holds the k-th pupa of each treatment.
Vial_Ctrl and
Vial_Rapa (obsolete) give the batch and the vial of each pupa (e.g. fDEV009_54). The pupae of all vials and batches of a line are pooled, as the authors did. DGRPool's line means, medians and SDs reproduce the authors' summary file and Table S1 exactly for all 280 line x treatment combinations, with the same numbers of pupae (3 to 1,158 per line and treatment).
Summary dataset (140 lines):
-
mn_RapaDelay: rapamycin-induced developmental delay (mean pupation time on rapamycin - mean on control, in days; Table S1 "Pupation time difference"). It is computed here at full precision from the authors' line means and rounds to Table S1 for all 140 lines. It can be derived from raw_1 as mean(PupTime_Rapa) - mean(PupTime_Ctrl).
-
se_RapaDelay: its standard error (Table S1 "SE"). It equals sqrt(SEM
rapa2 + SEM
control2), not the pooled SE described in the Methods.
-
mn_RapaDelayGWAS: the phenotype of the authors' GWAS and heritability analyses (
GWASinput_FINAL.txt). The delay + 1 was Box-Cox transformed, then centred within each batch on the mean of the four batch-control lines and scaled by the batch SD, then averaged over batches. The batch statistics included the removed stocks. The authors' models added Wolbachia status and five inversions as covariates.
raw_2 (food dilution, one row per pupa; 2023 experiment of Figure 4c, not part of the screen): pupation time on food with the yeast and sugar at 100%, 75% or 50% of the standard amounts, with or without rapamycin (
PupTime_Food100_Ctrl ...
PupTime_Food50_Rapa). Up to three vials per line and condition, plus obsolete vial numbers (
Vial_Food100_Ctrl ...). Time is in days from noon of the egg-collection day (21 March 2023), as in the authors' script, with the census dates and times taken from the file header. The authors' script numbered the last four censuses (10, 12, 14 and 18 April) one day too early (days 19, 21, 23 and 27 instead of 20, 22, 24 and 28, 91 pupae). The calendar days are used here. Because the time origin (noon) differs, these times are not directly comparable with raw_1 (origin midnight). The dataset has all 17 lines of the authors' file.
Caution: DGRP_439 and DGRP_820 are kept as the authors used them (Figure 4c, as sensitive lines), but these stocks failed the authors' PCR genotype check in the 2019-2020 screen (so they are not in raw_1 or the summary dataset), and nothing says whether the 2023 stocks were checked again: they may not be the expected DGRP lines (250 and 335 pupae). The other 15 lines are in the screen. The paper reports 6 or 5 sensitive lines (text vs legend) and 7 resistant lines for this experiment; the authors' script drops lines with fewer than 10 pupae in a treatment.
raw_3 (larval metabolome, Figures 3-4): 154 aqueous metabolites (targeted LC-MS) in larvae of 13 lines (6 resistant, 7 sensitive), 2 days after egg laying on control (
_Ctrl) or rapamycin (
_Rapa) food. Each sample is a pool of up to 50 larvae from three vials. Values are the natural log of the peak intensity, centred and scaled within each sample over the 154 metabolites (the authors' normalisation). The authors then also removed the effect of the two extraction batches with ComBat; that step is not applied here. The batch is given in the obsolete column
MetBatch (1 = larvae collected on 14 October 2021, 2 = 15 October 2021). One row per line and collection day (21 rows; the control and rapamycin samples of a row were collected the same day), with the number of larvae per sample in
NLarvae_Ctrl and
NLarvae_Rapa (obsolete). As the authors did, a line and day were kept only if both samples had more than 20 larvae: 42 of the 52 samples were kept, and line 142 has no data. One excluded pair (line 318, first day) also had two samples labelled rapamycin. Within-line rapamycin - control differences correlate at r = 0.9998 (sensitive lines) and 0.997 (resistant lines) with the authors' effects in Table S4. As in the paper, the isotope-labelled internal standards and the compounds not detected in every sample (most of them were detected in none) are not included. Names are abbreviated (e.g. 3OH3MeGlutarate = 3-methyl-3-hydroxyglutaric acid); each description gives the full name.
Not available: the dose-response data (Figure 1b: 6 lines, 0-1,024 nmol rapamycin) and the larval sizes (Figures 1c and S2: 6 lines) are not in the repository; they are only shown in figures.
Notes on the source: 14 of the 1,110 screen vials produced more than 40 pupae (up to 49), although 40 embryos were placed per vial; kept as published. For one census in batch fDEV010 (28 August 2019, 5 pupae), the day number (19) is one more than the date implies; the authors' day number is kept. The Figure 1c legend gives delays of 0.02 d for Ral_535 and 5.5 d for Ral_348. In Table S1, these are the values of lines 176 (0.02) and 181 (5.54); lines 535 and 348 have -0.06 and 6.94.