Comments from curator
Phenotyping data comes from the
Supplemental Material of the paper (
journal page; full text also on
PMC), the authors' per-vial data files:
-
File S1 (FileS1.xls): development time, egg-to-adult viability and sex ratio of the 2,849 vials of the main experiment.
-
File S2 (FileS2.xls): development time of the 708 vials of the reciprocal-cross experiment.
-
File S3 (FileS3.xls): eggs laid per female, 2 egg-lay plates per genotype.
-
File S4 (FileS4.xls): coefficients of variation of the male and female development times, for the 2,399 vials of File S1 with at least 2 males and 2 females.
The supplement links of the journal no longer work (the old genetics.org links redirect, and the OUP site refuses automated downloads), so the four files were retrieved from the Internet Archive copies of the OUP supplement links (captured on 2022-12-20). They are the authors' original files (saved on 2016-03-07).
These are not the original DGRP lines: each of 12 DGRP lines (DGRP_304, 313, 315, 358, 375, 517, 707, 712, 714, 765, 786 and 820) received each of 6 foreign mitochondrial genomes by balancer chromosome substitution, followed by several generations of backcrossing to males of the original DGRP line (crossing scheme of Zhu
et al. 2014):
D. melanogaster OreR, AutW132 and Zim53,
D. simulans siI and sm21, and
D. mauritiana mau12. The 72 resulting mitonuclear genotypes have the nuclear genome of the DGRP line and a mtDNA that is not its own (the native DGRP mtDNAs were not included), and they were cleared of
Wolbachia with tetracycline. Each value is stored under the DGRP line of the nuclear genome, and the mtDNA haplotype is given in the phenotype names (e.g. DevTime_siI_HPC), as is done for F1 crosses of DGRP lines.
Diets (name suffix; composition per 200 mL water, Table 1 of the paper):
-
Lab: standard laboratory food (1.8 g agar, 5 g yeast, 10.4 g cornmeal, 22 g sucrose), first experimental block (April 2014), 6 vials per genotype.
-
HPC,
EPC and
LPC: three isocaloric diets with a high, equal or low protein:carbohydrate ratio (Matzkin
et al. 2011: 32, 20 or 8 g yeast and 8, 20 or 32 g sucrose, with 9 g cornmeal and 1 g agar), second block (December 2014), 12 vials per genotype.
The laboratory food is therefore confounded with the block. The sm21 mtDNA on the DGRP_313 background was not tested on the three isocaloric diets, as stated in the paper.
Main experiment (dataset 1, Files S1 and S4): about 50 mated flies of each genotype laid eggs for 12 h on grape juice-agar plates, and up to 30 eggs were placed in each vial of 10 mL food (fewer in 630 of the 2,849 vials, as in the paper), kept at 25 °C (12:12 h light:dark). Eclosing adults were counted and sexed twice a day (9 am and 5 pm). The unit (row) is the vial, since the files give no per-fly times. For each mtDNA (MT) and diet (DIET):
-
DevTime_MT_DIET: egg-to-adult development time (hours), mean of the females (sex F), of the males (sex M) or of all adults of the vial (sex NA).
-
Viab_MT_DIET: egg-to-adult viability (adults / eggs).
-
SexRatio_MT_DIET: proportion of males among the adults.
-
MFDiff_MT_DIET: male minus female mean development time of the vial (hours), the paper's DT
MALE − DT
FEMALE.
-
DevCV_MT_DIET: coefficient of variation of the development time among the males (sex M) or among the females (sex F) of the vial, from File S4.
-
Eggs_MT_DIET (obsolete design column): number of eggs placed in the vial.
On a row, the columns of one mtDNA × diet combination come from the same vial, but each combination is a separate set of vials: columns of different combinations on the same row do not come from the same vial. The paper's male–female sex bias (number of males − number of females) equals (2 × SexRatio − 1) × Viab × Eggs, and its CV
MALES − CV
FEMALES is the difference between the M and F DevCV values of a vial. The authors log-transformed the development times for their models. Not uploaded: the egg-lay plate IDs (a random effect in the authors' models), and the sums of development hours and counts of adults, which can be derived from the uploaded values.
Reciprocal crosses (dataset 2, File S2): in 7 backgrounds, flies carrying two different mtDNAs (with the same DGRP nuclear genome) were crossed in both directions, along with the two within-genotype crosses, and the development time of the offspring was scored on the laboratory food.
DevRC_X is the within-genotype cross (X × X), and
DevRC_XxY is the cross of X females with Y males, whose offspring carry the mother's X mtDNA. Each value is the mean development time (hours) of the flies of one vial (both sexes, sex NA). The number of eggs per vial (30 for nearly all vials, missing for DGRP_315) is not uploaded. In File S2, 24 vials of DGRP_375 are labelled "mau12 x mau13" to "mau12 x mau36": they are vials 2 to 25 of the "mau12 x mau12" cross (whose vial 1 is labelled correctly), renamed by a spreadsheet auto-increment, and they were uploaded as DevRC_mau12.
Fecundity (dataset 3, File S3):
EggsFem_MT is the number of eggs laid per female on one egg-lay plate during the 12-h egg lay of the first block (total eggs / number of females, 13 to 30 females per plate), 2 plates per genotype. For DGRP_315 with the mau12 mtDNA, the two plates have identical values (492 eggs, 29 females), possibly a duplicated row.
Checks against the paper: the files reproduce the number of vials (2,849, of which 630 had fewer than 30 eggs: mean 12.23 ± 6.87, range 1–29), the range of the vial means (192–288 h), the missing sm21;DGRP-313 diets, the correlation between male and female vial means (r = 0.90, d.f. = 2,397, vials with at least 2 males and 2 females), the eggs-per-female ANOVA of Table 2F (identical sums of squares) and all the correlations of Figure 6 (CV of the male and female vial means within each group). The counts of adults in File S1 (19,071 males and 19,193 females) are about 1% lower than in the paper (19,274 and 19,408), and the global male delay is 2.10% instead of 2.07%.
Of note:
- For 5 vials of the laboratory food (DGRP_358 with mau12, DGRP_375 with siI, DGRP_517 with OreR, DGRP_712 and DGRP_820 with sm21), the mean development time of all adults (sex NA) is lower than both the male and the female means of the same vial, which is impossible. The values are kept as published.
- The column of summed development hours of File S1 (not uploaded) is wrong for 330 vials of DGRP_313 and DGRP_315 on the isocaloric diets, but their mean development times are consistent with the male and female hours.
- File S1 has no vial without adults (the lowest viability is 1/30), and 18 vials received 31 to 40 eggs.