Comments from curator
Phenotyping data comes from the
Supporting Information of the paper (
journal page, also on
PMC):
-
S2 Table (pone.0302565.s005.xlsx): FC50 of each replicate assay, 79 replicates of 31 DGRP lines.
-
S3 Table (pone.0302565.s006.xlsx): mean, standard deviation and number of replicates (N) of FC50 per line.
These are F1 phenotypes, not phenotypes of the inbred DGRP lines: virgin females of the
AKHRi model strain (
w; UAS-AKHR-RNAi; r4-GAL4, homozygous;
r4-GAL4 outcrossed to
w1118; BDSC 33832 and 51710) were crossed to males of each DGRP line. The F1 larvae knock down the adipokinetic hormone receptor
AKHR (the fly glucagon receptor) in the fat body, which makes them fatter and more buoyant, so their density measures the severity of this obesity model in a heterozygous DGRP background. Each value is assigned to the DGRP father line, as in the authors' GWAS, and the cross is given in the phenotype names (_xAKHRi). Larvae were collected at the L1 stage and raised at room temperature on a control medium (6% yeast, 6% dextrose, 3% sucrose, 1% agar) at 75-100 larvae per vial until the wandering L3 stage. Larvae were not sexed, so the sex is NA.
Density assay: 30 wandering L3 larvae, cleaned on damp filter paper, were placed in 10 mL of 8% sucrose; 250 µL of 20% sucrose was added repeatedly and the floating larvae scored after 2 min at each step, until 95-100% of them floated.
FC50 is the sucrose concentration (in % sucrose, unit %) at which 50% of the larvae floated: a lower FC50 means more buoyant, fatter larvae (a proxy for fat content).
Raw dataset (S2 Table):
FC50_xAKHRi has one row per replicate assay of 30 larvae (79 replicates, 1-4 per line). The Rank column of S2 Table (rank of the line mean) is not uploaded. A one-way ANOVA of FC50 on the line with these 79 values gives P = 0.0385, as in the paper (Fig 1).
Summary dataset (S3 Table):
mn_FC50_xAKHRiAuth and
sd_FC50_xAKHRiAuth, the authors' mean and standard deviation per line (SD NA for the 5 lines with a single replicate). The mean is the phenotype of the authors' preliminary GWAS (GEMMA; 213 variants at P < 10
-4 in 84 candidate genes). The line means and SDs derived by DGRPool from the raw dataset reproduce S3 Table for 30 of the 31 lines, but not for DGRP_304 (see below), which is why the authors' values are uploaded too. Line means range from 10.6875% (DGRP_555) to 12.845% sucrose (DGRP_307).
Of note:- DGRP_304: S3 Table gives N = 4, mean 11.8925 and SD 0.689, but S2 Table has only 3 replicates (11.65, 11.31, 11.72; mean 11.56, SD 0.219). The S3 values imply a 4th replicate of 12.89, which is not in S2 Table and is not uploaded; the paper's ANOVA P-value is reproduced without it (P = 0.0527 with it).
- DGRP_021: S3 Table gives N = 3, but its mean and SD (12.115 and 0.049) are those of the 2 replicates of S2 Table.
- The paper reports 81 assays (2,430 larvae), the sum of the N column of S3 Table; S2 Table has 79.
- The GWAS model in the Methods is described with "average glucose concentration" (copied from the authors'
Sirt1 screen,
DGRPool study 201); the GWAS phenotype is the average FC50.
Not uploaded (not DGRP lines): the validation of the model against its genetically matched control (S1 Fig), the density assays of the candidate modifier RNAi and overexpression strains crossed to the model or to
r4-GAL4 (Fig 2, S3 Fig, S6 Table), their triglyceride, glucose, glycogen and protein levels on control, high-sugar and high-fat diets (Figs 3-7) and the qPCR validation (S2 Fig). Also not uploaded: the qPCR primers (S1 Table) and the GWAS results (S4 and S5 Tables).