Embryogenesis is an essential and stereotypic process that nevertheless evolves among species. Its essentiality may favor the accumulation of cryptic genetic variation (CGV) that has no effect in the wild-type but that enhances or suppresses the effects of rare disruptions to gene function. Here, we adapted a classical modifier screen to interrogate the alleles segregating in natural populations of C. elegans: we induced gene knockdowns and used quantitative genetic methodology to examine how segregating variants modify the penetrance of embryonic lethality. Each perturbation revealed CGV, indicating that wild-type genomes harbor myriad genetic modifiers that may have little effect individually but which in aggregate can dramatically influence penetrance. Phenotypes were mediated by many modifiers, indicating high polygenicity, but the alleles tend to act very specifically, indicating low pleiotropy. Our findings demonstrate the extent of conditional functionality in complex trait architecture.
Embryonic lethality data
This is a tab-delimited text file with one header row, 27,360 data rows, and 19 columns. Each data row corresponds to a unique image taken of a well in a 96-well plate, in which there are (potentially) adult worms, dead embryos, and living larvae. The column headers are: adults_man (the number of adults in the image, manually counted by a human); eggs_man (the score of the abundance of dead embryos in the image, from 0 to 9, scored manually by a human); larvae_man (the score of the abundance of living larvae in the image, from 0 to 9, scored manually by a human); strain (the worm strain); vector (the gene targeted by RNAi, or in the case of "empty," the negative control RNAi clone); date (the one of three date batches in which the experiments were conducted); well_row (the row, from A to H, in a 96-well plate, in which the well was located); well_col (the column, from 1 to 12, in a 96-well plate, in which the well was located); fol_strain (the grouping, from a to c, to which the worm strain in this image belonged when it was dispensed into the 96-well plates; strains were dispensed in sets of 8, as each plate has 8 rows); fol_lib (the RNAi library, from 1 to 4, to which the bacteria in this well belonged when it was dispensed); fol_bac_rep (the batch of bacterial culture, either "car," "gre," or "roy" that was dispensed into this well; from a single bacterial library, three replicate culture batches were grown in three deep-well 96-well plates, from which the RNAi bacteria were dispensed into the experimental plates); fol_plate_rep (the replicate plate number, from 1 to 8; each experimental condition of worm strain and RNAi vector was replicated 8 times, across the three replicate batches of bacterial culture); image (the image number, from 1 to 96, in a 96-well plate, assigned by the image capture software); adult_area (the number of pixels assigned to the "adult" category by the DevStaR algorithm); larva_area (the number of pixels assigned to the "larva" category by the DevStaR algorithm); egg_area (the number of pixels assigned to the "embryo" category by the DevStaR algorithm); adults (the count of adults in the image by the DevStaR algorithm); larvae (the count of living larvae in the image by the DevStaR algorithm); eggs (the count of dead embryos in the image, determined by dividing egg_area by 70). NA values in the manually scored columns (adults_man, eggs_man, larvae_man) indicate that the well failed to grow a mature culture of worms. This is often due to somatic lethality of the targeted gene (e.g. the somatic control vector tba-2), or to fungal contamination. Because these images were checked by a human eye, rows with NAs in the manual columns should be excluded from most analyses.
emb_leth_data.txt