Comments from curator
Phenotyping data comes from
Supplementary Data 2 of the paper (
Supplementary information, file 41559_2024_2587_MOESM4_ESM.xlsx, sheet DGRP_phenotypes; author manuscript on
PMC): mean and standard error of the sperm length and of the seminal receptacle (SR) length for 126 DGRP lines, identified by their Bloomington stock number, DGRP name and RAL name (which all agree). Only these per-line values were published (no per-fly data), so the study has a summary dataset only.
The 126 lines (from the Bloomington stock center) were reared at a controlled density: 70 first-instar larvae per vial with 8 mL of cornmeal medium, at 22.5 °C. Adults were collected on the day of eclosion, kept in single-sex groups of 10 with live yeast for 5 days, then dissected. For each male, 7 sperm were measured, and the male's sperm length is the mean of the 5 longest (to avoid underestimates from broken sperm); 12 males per line (1,512 males, 7,560 sperm). The SR, the primary sperm-storage organ of
Drosophila females, was measured in 16 females per line (2,016 females). Measurement methods are those of
Syed et al. 2021. SR length quantifies the female preference for long sperm: longer sperm are better at displacing, and resisting displacement by, shorter sperm in the SR, and this advantage increases with SR length.
The summary dataset has:
-
mn_SpermLength and
se_SpermLength: mean and standard error of the sperm length (mm), sex M.
-
mn_SemRecepLength and
se_SemRecepLength: mean and standard error of the SR length (mm), sex F.
Values are uploaded as stored in the file (the spreadsheet displays 3 decimals but holds full precision). The interline coefficients of variation computed from these means reproduce the paper (6.96% for sperm length, 9.75% for SR length), and the two traits are not correlated among lines (r = 0.03, P = 0.71), as in Fig. 2a. Line means range from 1.635 mm (DGRP_365) to 2.585 mm (DGRP_075) for sperm length and from 1.811 mm (DGRP_304) to 3.063 mm (DGRP_373) for SR length.
GWAS phenotype: the authors ran separate GWAS of the line means (DGRP2 web server, 1,884,703 variants with MAF ≥ 0.05, adjusted for inversions and
Wolbachia). For sperm length, the 4 lines with the longest sperm (Tukey outliers: DGRP_075, DGRP_109, DGRP_149 and DGRP_083) were first Winsorized to the mean of the 5th longest line (DGRP_195, 2.16223636363636 mm). Supplementary Data 2, and therefore mn_SpermLength, holds the non-Winsorized means (the published CV of 6.96% is the non-Winsorized one). To reproduce the GWAS phenotype, set mn_SpermLength of these 4 lines to the value of DGRP_195.
Of note: 13 lines (DGRP_073, DGRP_075, DGRP_083, DGRP_088, DGRP_109, DGRP_149, DGRP_153, DGRP_161, DGRP_176, DGRP_195, DGRP_309, DGRP_317 and DGRP_375) have a sperm length standard error of 0.0008-0.0013 mm, about ten times smaller than all other lines (0.0041-0.026 mm). They include the 5 lines with the longest sperm. These values are uploaded as published, but they may not have been computed like the others. The broad-sense heritabilities of the paper (0.69 for sperm length, 0.67 for SR length) were estimated from the individual data and cannot be recomputed from the published values.
Not uploaded: body size (thorax length), which was measured on the DGRP lines with the same methods but is not published for them; Supplementary Data 1 (sperm, SR and thorax length of 149 drosophilid species, not DGRP lines); Supplementary Data 3-6 (GWAS variants, candidate genes, Gene Ontology and molecular evolution analyses).