Comments from curator
Phenotyping data comes from
S1 Data ("Data and R code for all statistical analyses and to generate all figures", a zip), which was missing from the original article and was published with its
correction (
direct link). Files of its "Data and Code" folder used:
-
Lethal Load and Lifespan.csv: one row per fly of the first experiment (lifespan following infection and viral load at death), used for raw dataset 1.
-
DCV Load and Shed Titres.csv: one row per fly of the second experiment (viral load and virus shedding 1, 2 or 3 days post-infection), used for raw dataset 2.
-
DCV Load and Shed Proportions.csv,
V Outbreak Likelihood.csv and
Vsummarytable.csv: the authors' values per line and sex plotted in Figs 2A, 3A, 4B and 4C, used for the summary dataset.
Ten DGRP lines (5 of the most resistant and 5 of the most susceptible to DCV in
Magwire et al. 2012) were infected by pricking 3-5-day-old flies in the pleural suture with a pin dipped in Drosophila C Virus (DCV). Males and mated females (kept 24 h as a male-female pair) are the default condition. Virgin females are separate phenotypes with the suffix
_Virg (sex F). Each row of the raw datasets is one fly.
Raw dataset 1 (540 flies, 7-20 per line and group, as in S1 Table):
LifespanDCV is the number of days from infection to death (flies checked daily).
LogVLAD is the viral load of the same fly at death, as log10(DCV RNA copies per fly + 1) (the authors' column, copied as is; 0 = no DCV detected). InfectionDate (two infection days) and FlyID are kept as obsolete design columns.
Raw dataset 2 (999 flies): each fly was kept alone in a tube with food and sampled once, 1, 2 or 3 days post-infection (suffix _D1, _D2, _D3).
LogDCVLoad is the DCV load of the fly.
LogDCVShed is the DCV shed into its tube during the previous 24 h. Both are log10(copies + 1), computed as in the authors' R script: log10(Viral Load × 375 + 1) and log10(Virus Shed × 250 + 1). Zeros (no DCV detected: 439 loads and 501 sheddings) are kept, as in the authors' analyses. The source has no load value for 150 flies and no shedding value for 39 flies (NA). Two flies with neither value and 57 empty rows were dropped. InfectionBlock (infection day 1-3) and SampleID are kept as obsolete design columns.
The authors analysed load and shedding in two parts: the proportion of flies with detectable DCV (Figs 2A and 3A), and the mean of the non-zero values only (Figs 2B and 3B). The means computed by DGRPool from raw dataset 2 include the zeros, so they match neither; the authors' means are obtained by leaving out the zeros. The authors' proportions are in the summary dataset (
mn_PropLoad_* and
mn_PropShed_*, kept as published). For 4 of the 87 line/sex/day groups, the counts behind them differ from the per-fly data:
- DGRP_059 mated females, 2 DPI: load detected in 8 of 19 flies, vs 8 of 11 in the per-fly data.
- DGRP_059 males, 2 DPI: shedding in 2 of 12 flies, vs 10 of 12 (apparently swapped).
- DGRP_075 males, 1 DPI: load in 10 of 11 flies, vs 9 of 10.
- DGRP_818 males, 3 DPI: shedding in 7 of 8 flies, vs 6 of 7.
mn_PropV and
mn_VrelNonZero come from the authors' simulation of individual transmission potential, V = log10(virus shed + 1) × lifespan / nearest-neighbour distance. For each line and sex (males and mated females), 60 individuals were simulated by sampling the three traits independently: shedding and lifespan from this study, and nearest-neighbour distance from the social aggregation data of
DGRPool study 121 (Siva-Jothy & Vale 2019). mn_PropV is the proportion of individuals with V > 0 (Fig 4B), and mn_VrelNonZero is the mean V/V
mean of those with V > 0 (Fig 4C). These are model outputs, not measurements.
Mismatches with the paper: S2 Table gives the number of flies per line, sex and day; 81 of its 87 counts equal the rows of the titres file, and the other 6 differ by one fly. The text gives a mean lifespan of 11.4 days for RAL-765 females, whereas the data give 11.47. The other lifespans quoted in the text are reproduced: males 14.15 days (text 14.1; RAL-75 16.0, RAL-765 11.1) and mated females 13.07 (text 13.06; RAL-59 15.1).
Not uploaded: the per-fly qPCR Ct values and quantities; the "Infectotype" class (S1 Fig); the
pastrel SNP genotype of each line (a DGRP annotation); and the "Rounded Viral Load" columns, which are not aligned with the rows of the titres file and are not used by the authors' script.