Comments from curator
Phenotyping data comes from the authors' data repository, given in the paper's Data Availability statement (
GitHub DworkinLab/ChandlerChariPLoSGenetics2017, archived on figshare as
10.6084/m9.figshare.5579296):
-
Data/GS_Wings_RawArea_Oct2015.csv: wing area and semi-quantitative wing score of each measured fly of the DGRP experiment (Supplementary Fig S3), used for the raw dataset. Its design is described in Data/README_expt_severity_test.txt, and the authors' analysis in Scripts/FigS3_GB_severity_dependence_model_test_script.R.
The main experiments of the paper (allelic series and complementation of
scalloped (
sd) and
vestigial (
vg) alleles introgressed into the Samarkand and Oregon-R wild-type strains, rearing temperature, imaginal disc stainings) do not use DGRP lines and are not uploaded.
These are F1 phenotypes, not phenotypes of the inbred DGRP lines: females of the Samarkand (SAM) wild-type strain carrying one of three X-linked
sd alleles introgressed into SAM (
sd1, X-ray induced;
sdE3, P{wE} insertion;
sd58d, gamma-ray induced, re-introgressed into SAM from Oregon-R for this experiment), or wild-type SAM females, were crossed to males of each DGRP line and the progeny reared at 24 °C. The F1 males are hemizygous for the
sd allele (their X chromosome comes from the SAM mother) and heterozygous SAM/DGRP for chromosomes 2, 3 and 4, so recessive and X-linked modifiers of the DGRP lines are not detected. They were stored in 70% ethanol, and one wing per male was dissected and imaged (Leica M125 microscope, 50X, Leica DFC400 camera). Each value is assigned to the DGRP father line, and the cross is given in the phenotype names:
_xSAM (wild-type control cross),
_xSAMsd1,
_xSAMsdE3 and
_xSAMsd58d.
The raw dataset (raw_1) has one row per F1 male (one wing), sex M, 5 or 6 males per line and cross (2 for the
sdE3 cross of DGRP_341):
-
WingArea_*: wing area (column TotalArea), measured with the authors' ImageJ macro. The paper gives no unit or scale: the values (about 47 for wild-type F1 males, 7 for
sd58d) are not mm² and, despite the label "Wing Size (in pixels)" in the authors' script, not pixel counts, so no unit is set and the values are comparable within this study only.
-
WingScore_*: semi-quantitative ordinal score of wing morphology (column Semiqt_Num), from 10 (wild type) to 1 (most severe); here 10 to 2 (letters A to I in column Semiqt_Raw). All wild-type (_xSAM) F1 males score 10.
The four crosses are different flies, so they are stacked compactly: row k of a line holds the k-th male of each cross, and the area and score of a male are on the same row. The authors analysed the means per line and cross (raw and log-transformed wing area, score) and the among-line variability per allele (Levene's statistic); there was no GWAS.
Lines: the paper says the
sd alleles were crossed to 16 DGRP lines, but the data file has 14 (RAL-304, 313, 315, 324, 341, 357, 358, 360, 362, 365, 375, 379, 380 and 399). RAL-362 is not uploaded (8 males,
sdE3 and
sd58d crosses only), as the authors removed it from their analysis because the cross was potentially contaminated ("362 crosses were discarded due to experimental errors").
The dataset therefore has 13 DGRP lines and 300 males, the 13 lines of the authors' Fig S3. DGRP_341 (RAL-341) is not in DGRP Freeze 2 and was added to DGRPool for this study (not genotyped); the DGRP3 line RP_0341 is a different line.