Comments from curator
Phenotyping data comes from the
Supplementary Material of the paper (
journal page, also on
PMC; file jkag060_supplementary_data.zip):
-
Data S1 (Data_S1_G3-2026-406560.xlsx): expression of 9 antimicrobial peptide (AMP) genes, one sheet per gene, with the 3 individual samples of each line and condition and their averages.
-
Data S2 (Data_S2_G3-2026-406560.xlsx): percent early mortality of each injured vial, one sheet per diet and age, with the line averages.
Both are uploaded as raw data (one row per vial or per sample); DGRPool computes the line means. Line names (DGRP26, ...) were mapped to DGRP_026, ...
Design: 19 randomly chosen DGRP lines from the Bloomington stock center (DGRP_026, DGRP_073, DGRP_091, DGRP_093, DGRP_161, DGRP_352, DGRP_381, DGRP_382, DGRP_383, DGRP_385, DGRP_386, DGRP_409, DGRP_427, DGRP_439, DGRP_440, DGRP_441, DGRP_774, DGRP_892 and DGRP_897), reared at 25 °C on cornmeal-molasses-yeast food (CMYD). Traumatic brain injury (TBI) was inflicted with the high-impact trauma (HIT) device on vials of 60 flies (1:1 female:male) aged 0-7 days (young,
Yng) or 20-27 days (old,
Old): 4 strikes at 5-min intervals with the spring deflected to 90°. The flies were then kept at 25 °C in vials with CMYD (
CMYD) or with water only (filter paper soaked with 200 µL water,
Wat). The four age × diet conditions are in the phenotype names.
raw_1, early mortality (4 phenotypes MI24_<diet>_<age>, Data S2): percent of injured flies dead within 24 h minus percent of uninjured flies of the same line dead at the same time, which corrects for the intrinsic mortality of each line (mortality of uninjured flies was below 1%). This is the MI24 index of the lab's earlier papers, as in
DGRPool study 20 (Katzenberger et al. 2015); here it was newly measured on these 19 lines in four conditions. One value per injured vial: 1,397 vials, 10 to 26 per line and condition. The vials are mixed-sex (the authors note that females and males have equivalent early mortality), so the sex is NA. The vials of the four conditions are independent; they are stacked, the k-th row of a line holding the k-th vial of each condition (NA when a condition has fewer vials). The DGRPool line means equal the 'Average Mortality' rows of Data S2 (the line means of Fig. 3) for all 76 line × condition combinations.
raw_2, AMP expression (72 phenotypes <gene>_<diet>_<age>_<Uninj|Inj>, Data S1): expression of
AttC,
AttD,
CecC,
Def,
DiptB,
Dro,
Drs,
Mtk and
Mtkl by qRT-PCR of total RNA from 20 whole males (males only, to avoid sex differences in basal expression), normalized to
RpL32 (2^-ΔCt, unitless), in uninjured (Uninj) and injured (Inj) flies of each age × diet condition. Three biological replicates per line and condition (each run in technical duplicate; Data S1 gives one value per biological sample), one row each, sex M. The paper does not say how long after the injury the flies were collected for RNA. The k-th sample of every gene and condition is on the k-th row of the line: the samples are not labelled in Data S1, but their deviations from the condition mean are correlated across genes (median r = 0.18, vs 0.00 when shuffled), so the k-th samples of the different genes are most likely the same RNA sample.
The uploaded values reproduce the paper: for expression (log2), the ANOVA F values of injury, age and diet (F
1,4100 = 401.62, 107.14 and 10.23), the five-way interaction (F
144,2736 = 1.72, and 1.83 and 1.39 before and after injury), all of Table 1 and the four-way interactions of Fig. 2 (AttC, Def, Mtk), exactly; for early mortality (arcsine square root), the line × diet × age interaction (F
18,1321 = 8.56) and the per-line tests of DGRP_409, DGRP_381 and DGRP_440 exactly, and Table 2 within 0.8 percentage points. The average correlation of expression with mortality is 0.31 in young and -0.14 in old flies, as in the text, and 14 of the 16 correlations printed in Fig. 4 are reproduced (
Drs, old injured flies: -0.68 and -0.44 from Data S1, -0.69 and -0.45 in the figure).
Of note, Data S1 has a few errors, kept as published (the paper's statistics include them):
-
Def and DiptB: 11 line × condition cells have exactly the same three values in both sheets: the 8 conditions of DGRP_892, the two water/old conditions of DGRP_897 and CMYD/young/injured of DGRP_774. These values are typical of
DiptB (0.4 to 2.4 times the median of the 19 lines) but up to 43 times the
Def median (more than twice it in 8 of the 11 cells), so they are probably DiptB values copied into the Def sheet.
-
DiptB_Wat_Yng_Inj of DGRP_382: 2 of the 3 values (0.984 and 0.955) are spreadsheet formulas (=2^-C51, =2^-C52) that apply 2^-x to the uninjured value of the same row, so they are probably wrong. The cached values are uploaded.
-
Typed averages: the line averages of Data S1 (typed, rounded to about 4 significant digits; not uploaded) equal the mean of the three samples within 0.1% for 1,343 of the 1,368 line × condition combinations, but 13 differ by more than 1% (typed vs mean of the samples): AttD_Wat_Yng_Uninj of DGRP_381 (0.0001035 vs 0.0001096); AttD_CMYD_Yng_Uninj of DGRP_409 (0.0005084 vs 0.0007751); CecC_CMYD_Yng_Uninj of DGRP_381 (0.0004366 vs 0.0004507); CecC_CMYD_Yng_Uninj of DGRP_439 (5.18e-05 vs 0.0005182); CecC_Wat_Old_Uninj of DGRP_382 (0.0090526 vs 0.009526); CecC_Wat_Yng_Uninj of DGRP_892 (0.000484 vs 0.0005609); DiptB_Wat_Old_Uninj of DGRP_381 (0.0485215 vs 0.04244); DiptB_Wat_Yng_Inj of DGRP_382 (0.8547932 vs 0.9832); Dro_CMYD_Yng_Uninj of DGRP_352 (0.0002797 vs 0.0002599); Drs_Wat_Old_Uninj of DGRP_382 (0.175943 vs 0.1344); Drs_Wat_Old_Uninj of DGRP_897 (0.13444 vs 0.1558); Drs_Wat_Old_Inj of DGRP_897 (0.357853 vs 0.2641); Mtkl_Wat_Old_Inj of DGRP_093 (0.1336239 vs 0.8429).
In Data S2 (CMYD/old), DGRP_383 and DGRP_385 have the same values for their first 10 vials (38.33, 38.33, 28.33, 60, 60, 45, 53.33, 53.33, 60, 50), possibly a copy-paste error, kept as published. Data S2 values are given with varying precision (e.g. 8.3 and 8.33, 41.676), kept as published.
Not uploaded: Data S3-S6 and Figs. 6-8 (AMP expression, early mortality and lifespan of
Dif and
Rel deletion mutants,
Difdel and
Reldel, and their controls, which are not DGRP lines) and Table S1 (primers).