Study

Curated
Bak et al., 2025
The role of genetic variation in shaping phenotypic responses to diet in aging Drosophila melanogaster
[10.1038/s41437-025-00797-3]

Description

Lifespan, and locomotor activity, heat knockdown time and dry weight from 7 to 61 days of age, were measured on males of 98 DGRP lines fed a control or a restricted diet (25% nutrient dilution). All traits showed genotype-by-diet and genotype-by-age interactions; GWAS implicated PGRP-LC and mbl.

Comments from curator

Phenotyping data come from two sources. The individual-fly data (raw datasets raw_1 and raw_2) are from the authors' Dryad deposit (doi:10.5061/dryad.0zpc8679m, version of 25 September 2025: healthspan_data.xlsx and lifespan_data.xlsx). The published line means and standard errors (SE) (summary dataset) are from Supplementary material 2 of the paper (41437_2025_797_MOESM2_ESM.xlsx, Supplementary information): Table S12 (lifespan), Table S13 (locomotor activity), Table S10 (dry weight) and Table S14 (heat knockdown time, HKDT), per diet and, for the three healthspan traits, per age.

Design. Only males were tested (F2 males; females were not used because their virginity could not be ensured). At 2 days (± 36 h) of age, males of 98 DGRP lines were placed on either a control diet (CD, 478 cal/L: yeast 60 g/L, sucrose 40 g/L, oatmeal 30 g/L) or a restricted diet (RD, 119.5 cal/L: the same diet diluted to 25% of its nutrient content with α-cellulose), at 23 °C, in up to three bottles of 90 males per line and diet. The diets were given to adults only, not during development.
- Lifespan (days): dead flies were counted every 3 days, at each transfer to fresh food, until all flies had died.
- Healthspan traits, measured on the same individual flies, taken from the bottles at 7 days of age and then every 9 days (7, 16, 25, 34, 43, 52 and 61 days; suffix _7d to _61d), 8 flies per line and diet at 7 days and up to 16 later:
  - LocAct (counts per minute): locomotor activity of the fly in a Drosophila Activity Monitor (DAM2) over 6 h at 23 °C in constant light.
  - HKDT (minutes): the monitors were then heated to 39 °C, and the heat knockdown time is the time of the last recorded activity count.
  - DryWt (mg): the flies were then dried for 48 h at 60 °C and weighed individually to the nearest 10 µg.

Raw datasets (one fly per value, all males). To keep the datasets compact, the conditions are stacked: row k of a line holds the k-th fly of each age and diet. Values on the same row and condition belong to the same fly; values of different conditions on the same row come from different, unrelated flies.
- raw_1 (healthspan_data.xlsx, 6,338 flies of 98 lines): LocAct_, HKDT_ and DryWt_<age>_<diet>. LocAct and HKDT are NA where the authors' file has NA: when a fly had no activity count in either assay, it was assumed dead. HKDT is also missing for 1,975 flies that have an activity value (the paper's Table S3 shows the same low HKDT numbers, e.g. 225 flies at 16 days on the CD), and 5 flies have an HKDT but no activity value. HKDT = 0 (145 flies) means no activity count after the temperature increase. Dry weight is in mg: the Dryad README says micrograms, but the values (0.07-0.36) and Table S10 are in mg. Rep_<age>_<diet> (obsolete design column) is the bottle (A, B or C) the fly came from. The letters were assigned at random within each line and diet, so bottles with the same letter are unrelated across lines and diets; the authors fitted the bottle as a random effect. The DAM monitor and position of each fly are in the Dryad file but were not uploaded.
- raw_2 (lifespan_data.xlsx, 43,750 flies of 99 lines): Lifespan_CD and Lifespan_RD hold the age at death (days, 5 to 59). The 8,683 censored flies are NA in Lifespan, because the authors left them out of their line means (Table S12) and, as the per-fly data show, of their Cox model (Table 3). They are the 6,338 flies taken for the healthspan assays and 2,345 flies that escaped, got stuck in the food or were trapped by the stopper. Their age at censoring is in CensorAge_CD/RD (obsolete), and the bottle of every fly is in Rep_CD/RD (obsolete).

Comparison with the published line means. DGRPool's line means and SE (SD/√n) of the raw phenotypes equal the published values of Tables S10 and S12-S14 for 2,289 of the 2,303 line × trait × age × diet values (differences below 4e-14). This includes every dry weight and lifespan value. The other 14 values are locomotor activity and HKDT means that include flies missing from the Dryad file: 10 flies with an activity value, 5 of which also have an HKDT. The paper counts 5,641 flies for locomotor activity and 3,666 for HKDT, against 5,631 and 3,661 in the file (see also Table S2), and adding one or two flies reproduces each of the 14 published means and SE exactly. The Dryad file holds the 6,338 flies with a dry weight, so the missing flies probably had none. The affected lines are:
  - locomotor activity: 7 days CD (DGRP_088, DGRP_105, DGRP_757), 7 days RD (DGRP_805), 16 days RD (DGRP_714), 25 days CD (DGRP_356, DGRP_379, DGRP_385) and 34 days CD (DGRP_859). DGRP_379 has no activity value at 25 days in the Dryad file.
  - HKDT: 7 days CD (DGRP_105, DGRP_757), 7 days RD (DGRP_805), 16 days RD (DGRP_714) and 25 days CD (DGRP_385).
For these 9 trait × age × diet combinations, the published means and SE are kept active as the authors' values, renamed mn_<Raw>Auth and se_<Raw>Auth (e.g. mn_LocAct_7d_CDAuth). They are the values used in the paper's analyses. The other 57 summary phenotypes are obsolete, because DGRPool computes the same values from the raw data.

Ages used in the paper. Fewer flies were alive at older ages (for example only 2 lines have restricted-diet values at 43 days, none at 52 and 61 days), so the authors' heritabilities, correlations and GWAS use the 7- and 16-day data only. The data at 25 to 61 days are in the authors' files and are uploaded too, but they are not used in the main analyses of the paper.

Notes on the values.
- The published SE is missing when a line mean is based on a single fly, and is 0 when all flies of the line had the same value.
- Tables S10, S13 and S14 contain the same 98 lines, but Table S12 (lifespan) contains DGRP_461 instead of DGRP_101. The Dryad files explain this. DGRP_101 has only 15 flies, all taken at 7 days for the healthspan assays, so it has healthspan data at 7 days but no lifespan. DGRP_461 has a single fly (RD, dead at 5 days) and no healthspan data. The data therefore cover 99 lines (98 in raw_1, 99 in raw_2 and in the summary); the paper reports 98.
- Table 2 of the paper (average trait means per diet and age) averages individual flies, not line means: the per-fly data reproduce it.
- The Cox model of Table 3 (diet effect: β = 0.25859, SE 0.01117) is reproduced exactly from the 35,067 dead flies alone (Efron ties), although the text says that all 43,750 flies, including the censored ones, were used. With the censored flies, β is 0.528.
- Table S3 gives 228 HKDT flies at 25 days on the CD in the unbalanced data, which is fewer than its balanced subset (257). The Dryad file has 287 (288 with the missing fly), so 228 is probably a typo.
- 46 of the 539 bottles have more than 90 flies in the lifespan file (up to 144, DGRP_303 on the CD).
- Lifespans are short for DGRP males (line means of 5 to 40 days, about 20 on average). Some lines have healthspan data at ages beyond their mean lifespan, because flies were taken for the assays from the same bottles.

GWAS. The authors ran GWAS on the line means of each trait for each diet at 7 and 16 days and for lifespan, adjusting for Wolbachia and five major inversions. The genotype-by-diet (and genotype-by-age) interaction GWAS used the per-line difference between the two diets (or between 7 and 16 days). It can be computed from the line means, e.g. mean of DryWt_7d_CD − mean of DryWt_7d_RD (use the mn_…Auth phenotypes to reproduce the paper's values where they differ). The interaction for dry weight at 7 days mapped to PGRP-LC (Table S18) and the interaction for lifespan to mbl (Table S19).

Of note, the age-related trait data of this experiment are also used in a companion paper by the same authors, Bak et al., 2026 (Biogerontology), which studies the gut microbiome of young flies and whose data statement points to the same Dryad deposit.

Other external data available

10.5061/dryad.0zpc8679m Dryad deposit of the paper (healthspan_data.xlsx and lifespan_data.xlsx, per-fly lifespan, locomotor activity, HKDT and dry weight data; uploaded as raw_1 and raw_2)

Datasets

Dataset Phenotypes DGRP lines
Summary dataset 75 phenotypes
mn_HKDT_16d_RDAuth mn_HKDT_25d_CDAuth mn_HKDT_7d_CDAuth mn_HKDT_7d_RDAuth mn_LocAct_16d_RDAuth mn_LocAct_25d_CDAuth mn_LocAct_34d_CDAuth mn_LocAct_7d_CDAuth mn_LocAct_7d_RDAuth se_HKDT_16d_RDAuth se_HKDT_25d_CDAuth se_HKDT_7d_CDAuth se_HKDT_7d_RDAuth se_LocAct_16d_RDAuth se_LocAct_25d_CDAuth se_LocAct_34d_CDAuth se_LocAct_7d_CDAuth se_LocAct_7d_RDAuth
99 DGRP lines
DGRP_021
DGRP_026
DGRP_059
DGRP_075
DGRP_083
DGRP_088
DGRP_091
DGRP_101
DGRP_105
DGRP_109
DGRP_129
DGRP_136
DGRP_149
DGRP_161
DGRP_176
DGRP_177
DGRP_181
DGRP_195
DGRP_217
DGRP_228
DGRP_239
DGRP_280
DGRP_287
DGRP_301
DGRP_303
DGRP_309
DGRP_313
DGRP_317
DGRP_318
DGRP_332
DGRP_335
DGRP_336
DGRP_338
DGRP_340
DGRP_356
DGRP_357
DGRP_358
DGRP_359
DGRP_360
DGRP_361
DGRP_362
DGRP_365
DGRP_373
DGRP_374
DGRP_375
DGRP_377
DGRP_379
DGRP_380
DGRP_382
DGRP_383
DGRP_385
DGRP_386
DGRP_391
DGRP_392
DGRP_399
DGRP_426
DGRP_427
DGRP_437
DGRP_440
DGRP_441
DGRP_443
DGRP_461
DGRP_491
DGRP_492
DGRP_502
DGRP_508
DGRP_517
DGRP_531
DGRP_535
DGRP_555
DGRP_589
DGRP_595
DGRP_639
DGRP_703
DGRP_705
DGRP_707
DGRP_712
DGRP_714
DGRP_716
DGRP_721
DGRP_732
DGRP_738
DGRP_748
DGRP_757
DGRP_761
DGRP_765
DGRP_774
DGRP_776
DGRP_783
DGRP_786
DGRP_796
DGRP_799
DGRP_801
DGRP_802
DGRP_804
DGRP_805
DGRP_820
DGRP_852
DGRP_859
Download
Raw dataset #1 36 phenotypes
DryWt_16d_CD DryWt_16d_RD DryWt_25d_CD DryWt_25d_RD DryWt_34d_CD DryWt_34d_RD DryWt_43d_CD DryWt_43d_RD DryWt_52d_CD DryWt_61d_CD DryWt_7d_CD DryWt_7d_RD HKDT_16d_CD HKDT_16d_RD HKDT_25d_CD HKDT_25d_RD HKDT_34d_CD HKDT_34d_RD HKDT_43d_CD HKDT_43d_RD HKDT_52d_CD HKDT_61d_CD HKDT_7d_CD HKDT_7d_RD LocAct_16d_CD LocAct_16d_RD LocAct_25d_CD LocAct_25d_RD LocAct_34d_CD LocAct_34d_RD LocAct_43d_CD LocAct_43d_RD LocAct_52d_CD LocAct_61d_CD LocAct_7d_CD LocAct_7d_RD
98 DGRP lines
DGRP_021
DGRP_026
DGRP_059
DGRP_075
DGRP_083
DGRP_088
DGRP_091
DGRP_101
DGRP_105
DGRP_109
DGRP_129
DGRP_136
DGRP_149
DGRP_161
DGRP_176
DGRP_177
DGRP_181
DGRP_195
DGRP_217
DGRP_228
DGRP_239
DGRP_280
DGRP_287
DGRP_301
DGRP_303
DGRP_309
DGRP_313
DGRP_317
DGRP_318
DGRP_332
DGRP_335
DGRP_336
DGRP_338
DGRP_340
DGRP_356
DGRP_357
DGRP_358
DGRP_359
DGRP_360
DGRP_361
DGRP_362
DGRP_365
DGRP_373
DGRP_374
DGRP_375
DGRP_377
DGRP_379
DGRP_380
DGRP_382
DGRP_383
DGRP_385
DGRP_386
DGRP_391
DGRP_392
DGRP_399
DGRP_426
DGRP_427
DGRP_437
DGRP_440
DGRP_441
DGRP_443
DGRP_491
DGRP_492
DGRP_502
DGRP_508
DGRP_517
DGRP_531
DGRP_535
DGRP_555
DGRP_589
DGRP_595
DGRP_639
DGRP_703
DGRP_705
DGRP_707
DGRP_712
DGRP_714
DGRP_716
DGRP_721
DGRP_732
DGRP_738
DGRP_748
DGRP_757
DGRP_761
DGRP_765
DGRP_774
DGRP_776
DGRP_783
DGRP_786
DGRP_796
DGRP_799
DGRP_801
DGRP_802
DGRP_804
DGRP_805
DGRP_820
DGRP_852
DGRP_859
Download
Raw dataset #2 2 phenotypes
Lifespan_CD Lifespan_RD
99 DGRP lines
DGRP_021
DGRP_026
DGRP_059
DGRP_075
DGRP_083
DGRP_088
DGRP_091
DGRP_101
DGRP_105
DGRP_109
DGRP_129
DGRP_136
DGRP_149
DGRP_161
DGRP_176
DGRP_177
DGRP_181
DGRP_195
DGRP_217
DGRP_228
DGRP_239
DGRP_280
DGRP_287
DGRP_301
DGRP_303
DGRP_309
DGRP_313
DGRP_317
DGRP_318
DGRP_332
DGRP_335
DGRP_336
DGRP_338
DGRP_340
DGRP_356
DGRP_357
DGRP_358
DGRP_359
DGRP_360
DGRP_361
DGRP_362
DGRP_365
DGRP_373
DGRP_374
DGRP_375
DGRP_377
DGRP_379
DGRP_380
DGRP_382
DGRP_383
DGRP_385
DGRP_386
DGRP_391
DGRP_392
DGRP_399
DGRP_426
DGRP_427
DGRP_437
DGRP_440
DGRP_441
DGRP_443
DGRP_461
DGRP_491
DGRP_492
DGRP_502
DGRP_508
DGRP_517
DGRP_531
DGRP_535
DGRP_555
DGRP_589
DGRP_595
DGRP_639
DGRP_703
DGRP_705
DGRP_707
DGRP_712
DGRP_714
DGRP_716
DGRP_721
DGRP_732
DGRP_738
DGRP_748
DGRP_757
DGRP_761
DGRP_765
DGRP_774
DGRP_776
DGRP_783
DGRP_786
DGRP_796
DGRP_799
DGRP_801
DGRP_802
DGRP_804
DGRP_805
DGRP_820
DGRP_852
DGRP_859
Download
Back to studies