Study

Curated
Wong et al., 2023
Pleiotropic fitness effects across sexes and ages in the Drosophila genome and transcriptome
[10.1093/evolut/qpad163]

Description

Early- and late-life fitness was measured in 125 DGRP lines: larvae produced in 24 h by groups of 5 females, and share of larvae sired by groups of 5 males competing with GFP-marked males. Fitness was positively genetically correlated across sexes and ages, with few sexually antagonistic loci.

Comments from curator

Phenotyping data comes from the authors' GitHub repository lukeholman/fitnessGWAS (code and data of the paper, rendered at lukeholman.github.io/fitnessGWAS):
- data/input/female_fitness_CLEANED.csv and male_fitness_CLEANED.csv: the authors' cleaned raw data, one row per vial (also shown on their page Tables of the raw data and estimated line means), used for raw dataset 1.
- data/input/female_fitness.csv and male_fitness.csv: the original files, used only for the fly counts that the authors' cleaning dropped (obsolete columns of raw dataset 1).
- data/input/larva to adult survival.csv: a larva-to-adult assay that is not described in the paper, used for raw dataset 2.
- The Estimated line means table of the same page: the authors' line means, used for the summary dataset.

Design: fitness was measured in males and females of 125 DGRP lines in ten experimental blocks (A2-A4, B1-B3, C1-C3 and E1 in the data; the supplementary methods say nine blocks), with DGRP_352 included in every block as a reference (60 vials per sex), DGRP_361 and DGRP_882 in two blocks (12 vials per sex) and every other line in one block (6 vials per sex). All flies were 2- to 3-day-old virgins at the start, kept on cornmeal-dextrose-yeast-soy flour food (Table S1).
- Female fitness: 5 females of the DGRP line were housed with 15 GFP-marked males and 10 bw competitor females in a lightly yeasted food vial for 3 days, then moved alone to a grape juice agar vial for 24 h, and the first-instar (L1) larvae that hatched were counted (early life, females aged 5-6 days). After 8 more days (one transfer) and 3 days with 15 new virgin GFP males, a second 24 h egg collection was counted the same way (late life, females aged 17-18 days).
- Male fitness: 5 males of the DGRP line competed with 10 GFP males for 15 bw females for 3 days; the females then laid eggs for 24 h and up to 200 L1 larvae were scored for GFP (early life). The males were left to age in the same vial (one transfer) and, at 14-15 days of age, were given 15 new virgin bw females for 3 days, followed by another 24 h egg collection scored for GFP (late life).
Dead flies were not replaced. Male fitness was not measured for DGRP_354 (experimenter error, according to the authors).

Raw dataset 1 (fitness): one row per vial, i.e. per group of 5 DGRP flies. Female rows (sex F) and male rows (sex M) come from different vials.
- FemFit_Early and FemFit_Late: number of L1 larvae produced in 24 h by the 5 females, as in the authors' cleaned file. The authors set the missing values of the original file to 0 (1 early-life vial and 40 late-life vials, 25 of which had no female left alive) and used these zeros in their models, so they are kept.
- MaleFit_Early and MaleFit_Late: proportion of the scored larvae sired by the DGRP males, i.e. GFP-negative / (GFP-negative + GFP-positive) larvae, computed per vial as the authors did for their repeatability estimates (Table S2). The underlying counts are given in FocalLarv_* (GFP-negative, sired by the DGRP males) and RivalLarv_* (GFP-positive, sired by the GFP males). MaleFit is NA when no larva was scored (1 early-life and 47 late-life vials, whose counts are kept as 0) or when the counts are missing (2 early-life and 3 late-life vials).
- Obsolete columns: the experimental block (FitBlock), and five fly counts of the original files that the authors neither describe nor use (FemLaying_Early, FemAlive_Late1, FemAlive_Late2, MaleAlive_Late, GFPMaleAlive_Late; presumably the numbers of flies alive at each assay). Some of their values are impossible (6 females where 5 were set up, up to 15 GFP males where 10 were set up), and some are missing for whole blocks.

Summary dataset (mn_FemFit_EarlyAuth, mn_FemFit_LateAuth, mn_MaleFit_EarlyAuth, mn_MaleFit_LateAuth): the authors' line means, which they used for all the analyses of the paper (heritabilities aside): genetic correlations, GWAS (GEMMA) and TWAS. They come from Bayesian multivariate mixed models (brms) of the vial data with random effects of line, block and vial: Poisson models of the larva counts for females, and binomial models of the number of larvae sired by the DGRP males out of the larvae scored for males. The line effects were predicted on the scale of the linear predictor (log or logit) and then standardised to mean 0 and SD 1 across lines, so they are unitless z-scores and not means of the raw data. Values have 3 decimals, as published on the authors' page; the full-precision file is not in the repository. The plain line means that DGRPool computes from raw dataset 1 rank the lines similarly (Spearman correlation with the authors' values: 0.87 and 0.88 for female early and late fitness, 0.75 and 0.75 for male early and late fitness).
Inconsistency in the source: in the authors' line-mean script (get_predicted_line_means.Rmd), the 47 late-life male vials in which no larva was scored were entered as 100 larvae sired by the DGRP males out of 100, i.e. full paternity, whereas the code comment says that they were meant to get zero fitness because all 5 DGRP males had died (in 45 of these vials the original file records 1-5 DGRP males still alive). The published mn_MaleFit_LateAuth values reflect this coding: the line means of MaleFit_Late, on the logit scale, correlate with them at r = 0.81 when these vials are counted as 1, against 0.66 when they are left out (NA, as in raw dataset 1) and 0.65 when they are counted as 0. For early-life male fitness, the authors imputed the vial without larvae and the 2 vials with missing counts (mice, 5 imputed datasets).

Raw dataset 2 (larva-to-adult survival, not described in or used by the paper): the repository file 'larva to adult survival.csv' gives, per vial, the number of L1 larvae placed (usually 100), the numbers of adult males and females that eclosed, the number of dead pupae and the times (h) to the first and last emergence. It covers 139 DGRP lines (23 of them are not in the fitness data) in ten blocks with the same names as the fitness blocks (A1-A4, B1-B3, C1-C3), usually 2 vials per line and block (29 vials for DGRP_352). The authors do not say where the larvae came from (presumably from the stocks of each DGRP line), nor when the emergence times start (presumably from the transfer of the L1 larvae).
- Rows with sex NA hold the vial-level values: L2A_Viability (adults eclosed / L1 larvae placed, computed per vial), L2A_DeadPupae, L2A_FirstEclosion and L2A_LastEclosion. The emergence times are check times, and the last emergence time is often the last check of the block (e.g. 283.42 h for every vial of block A2), so it mostly reflects the end of the observation.
- Rows with sex M and F hold L2A_Eclosed, the number of adults of that sex that eclosed from the vial.
- The number of larvae placed (L2A_L1Seeded) and the block (L2A_Block) are obsolete design columns.
Four vials have more adults than larvae placed (viability 1.02 to 1.12): they are kept as published. One vial of line "230" (block C1) was dropped: RAL-230 is not in DGRP Freeze 2 and has no DGRP_ line in DGRPool. It may also be a mislabelled second vial of DGRP_229, the only other line with a single vial in that block.

The transcriptome analyses used the published DGRP expression data of Huang et al. 2015 (PNAS); no new molecular data were generated.

Datasets

Dataset Phenotypes DGRP lines
Summary dataset 4 phenotypes
mn_FemFit_EarlyAuth mn_FemFit_LateAuth mn_MaleFit_EarlyAuth mn_MaleFit_LateAuth
125 DGRP lines
DGRP_021
DGRP_026
DGRP_032
DGRP_040
DGRP_045
DGRP_057
DGRP_059
DGRP_075
DGRP_091
DGRP_093
DGRP_109
DGRP_129
DGRP_136
DGRP_138
DGRP_142
DGRP_153
DGRP_158
DGRP_161
DGRP_176
DGRP_181
DGRP_195
DGRP_208
DGRP_217
DGRP_228
DGRP_229
DGRP_235
DGRP_237
DGRP_239
DGRP_280
DGRP_301
DGRP_303
DGRP_306
DGRP_307
DGRP_309
DGRP_313
DGRP_319
DGRP_320
DGRP_321
DGRP_324
DGRP_332
DGRP_338
DGRP_350
DGRP_352
DGRP_354
DGRP_356
DGRP_357
DGRP_360
DGRP_361
DGRP_362
DGRP_367
DGRP_370
DGRP_371
DGRP_373
DGRP_375
DGRP_379
DGRP_380
DGRP_382
DGRP_383
DGRP_385
DGRP_391
DGRP_392
DGRP_395
DGRP_397
DGRP_399
DGRP_405
DGRP_406
DGRP_427
DGRP_437
DGRP_439
DGRP_440
DGRP_441
DGRP_443
DGRP_461
DGRP_486
DGRP_491
DGRP_502
DGRP_509
DGRP_513
DGRP_517
DGRP_530
DGRP_531
DGRP_535
DGRP_584
DGRP_589
DGRP_596
DGRP_703
DGRP_705
DGRP_707
DGRP_712
DGRP_716
DGRP_721
DGRP_732
DGRP_738
DGRP_748
DGRP_757
DGRP_761
DGRP_765
DGRP_776
DGRP_783
DGRP_787
DGRP_790
DGRP_796
DGRP_801
DGRP_802
DGRP_804
DGRP_805
DGRP_808
DGRP_810
DGRP_820
DGRP_821
DGRP_832
DGRP_837
DGRP_843
DGRP_850
DGRP_852
DGRP_853
DGRP_859
DGRP_879
DGRP_882
DGRP_897
DGRP_900
DGRP_907
DGRP_908
DGRP_911
DGRP_913
Download
Raw dataset #1 8 phenotypes
FemFit_Early FemFit_Late FocalLarv_Early FocalLarv_Late MaleFit_Early MaleFit_Late RivalLarv_Early RivalLarv_Late
125 DGRP lines
DGRP_021
DGRP_026
DGRP_032
DGRP_040
DGRP_045
DGRP_057
DGRP_059
DGRP_075
DGRP_091
DGRP_093
DGRP_109
DGRP_129
DGRP_136
DGRP_138
DGRP_142
DGRP_153
DGRP_158
DGRP_161
DGRP_176
DGRP_181
DGRP_195
DGRP_208
DGRP_217
DGRP_228
DGRP_229
DGRP_235
DGRP_237
DGRP_239
DGRP_280
DGRP_301
DGRP_303
DGRP_306
DGRP_307
DGRP_309
DGRP_313
DGRP_319
DGRP_320
DGRP_321
DGRP_324
DGRP_332
DGRP_338
DGRP_350
DGRP_352
DGRP_354
DGRP_356
DGRP_357
DGRP_360
DGRP_361
DGRP_362
DGRP_367
DGRP_370
DGRP_371
DGRP_373
DGRP_375
DGRP_379
DGRP_380
DGRP_382
DGRP_383
DGRP_385
DGRP_391
DGRP_392
DGRP_395
DGRP_397
DGRP_399
DGRP_405
DGRP_406
DGRP_427
DGRP_437
DGRP_439
DGRP_440
DGRP_441
DGRP_443
DGRP_461
DGRP_486
DGRP_491
DGRP_502
DGRP_509
DGRP_513
DGRP_517
DGRP_530
DGRP_531
DGRP_535
DGRP_584
DGRP_589
DGRP_596
DGRP_703
DGRP_705
DGRP_707
DGRP_712
DGRP_716
DGRP_721
DGRP_732
DGRP_738
DGRP_748
DGRP_757
DGRP_761
DGRP_765
DGRP_776
DGRP_783
DGRP_787
DGRP_790
DGRP_796
DGRP_801
DGRP_802
DGRP_804
DGRP_805
DGRP_808
DGRP_810
DGRP_820
DGRP_821
DGRP_832
DGRP_837
DGRP_843
DGRP_850
DGRP_852
DGRP_853
DGRP_859
DGRP_879
DGRP_882
DGRP_897
DGRP_900
DGRP_907
DGRP_908
DGRP_911
DGRP_913
Download
Raw dataset #2 5 phenotypes
L2A_DeadPupae L2A_Eclosed L2A_FirstEclosion L2A_LastEclosion L2A_Viability
139 DGRP lines
DGRP_021
DGRP_026
DGRP_028
DGRP_038
DGRP_040
DGRP_045
DGRP_048
DGRP_057
DGRP_059
DGRP_075
DGRP_091
DGRP_093
DGRP_129
DGRP_136
DGRP_138
DGRP_142
DGRP_153
DGRP_176
DGRP_177
DGRP_181
DGRP_195
DGRP_208
DGRP_217
DGRP_228
DGRP_229
DGRP_235
DGRP_237
DGRP_239
DGRP_280
DGRP_287
DGRP_301
DGRP_303
DGRP_304
DGRP_306
DGRP_307
DGRP_309
DGRP_313
DGRP_315
DGRP_317
DGRP_318
DGRP_320
DGRP_321
DGRP_324
DGRP_332
DGRP_336
DGRP_338
DGRP_350
DGRP_352
DGRP_356
DGRP_357
DGRP_358
DGRP_359
DGRP_360
DGRP_361
DGRP_362
DGRP_365
DGRP_367
DGRP_370
DGRP_371
DGRP_373
DGRP_375
DGRP_377
DGRP_379
DGRP_380
DGRP_382
DGRP_383
DGRP_385
DGRP_386
DGRP_391
DGRP_392
DGRP_395
DGRP_397
DGRP_399
DGRP_405
DGRP_406
DGRP_427
DGRP_437
DGRP_439
DGRP_440
DGRP_441
DGRP_443
DGRP_486
DGRP_491
DGRP_502
DGRP_509
DGRP_513
DGRP_517
DGRP_530
DGRP_531
DGRP_535
DGRP_555
DGRP_559
DGRP_563
DGRP_584
DGRP_589
DGRP_591
DGRP_596
DGRP_703
DGRP_705
DGRP_707
DGRP_716
DGRP_721
DGRP_730
DGRP_732
DGRP_738
DGRP_748
DGRP_757
DGRP_761
DGRP_765
DGRP_776
DGRP_783
DGRP_786
DGRP_787
DGRP_790
DGRP_796
DGRP_801
DGRP_802
DGRP_804
DGRP_805
DGRP_808
DGRP_810
DGRP_820
DGRP_821
DGRP_832
DGRP_837
DGRP_843
DGRP_849
DGRP_850
DGRP_852
DGRP_853
DGRP_859
DGRP_861
DGRP_879
DGRP_882
DGRP_897
DGRP_900
DGRP_907
DGRP_908
DGRP_913
Download
Back to studies