Overview of the behavioral screen and assay workflow.

a) Schematic of the screening strategy. High-confidence ASD (Cat 1) SFARI genes were identified (207), filtered for those with a Drosophila ortholog (201), and their corresponding MiMIC lines were selected (80). Virgin females were crossed with wild-type or MiMIC+/- males, and behavioral phenotyping was performed on male and female offspring (68 viable lines after screening). b) Activity assay schematic. Individual flies were monitored for 48 h using the DAM2 system, generating 51 output parameters. PCA was applied to identify global differences. Individual parameters were plotted using rose plots with parameters ordered from most wake- to most sleep-related. c) Social spacing assay schematic. A total of 20 flies were dispensed in a triangular arena (four arenas per trial). After 20 min, an image was captured and analyzed using FlyFinder, a custom program that automatically detects each fly, records coordinates, and calculates multiple spatial parameters (number of flies within 4 body lengths, nearest 3 neighbors, nearest neighbor, and global median).

Genetic screen hits

Genetic screening in Drosophila identifies modulators of sleep and activity.

a,b) Log2fold change heatmaps and significance plots for all 68 viable screened genes in males (a) and females (b). For the top 6 PCs, 95% confidence intervals were computed for two comparisons: (i) each mutant genotype versus wild-type and (ii) each genotype versus the pooled mean of all other mutant lines. SGPVs were calculated for each comparison. Genes that met hit criteria (SGPV = 0) for the wild-type comparison alone are shown in gray, whereas genes that met both criteria are shown in red. Rows are ordered by the estimated effect size for PC5 and displayed in descending order (n=2-3 replicates per line with 4 flies per replicate). c,d) Bar graphs show the relative contribution of each behavioral parameter to the indicated PC, expressed as a percentage of the total contribution for that component, shown separately for males (c) and females (d). Color indicates the mean log2fold change between mutants and wild-type, with red denoting higher values in mutants and blue denoting higher values in wild-type. Y-axis labels denote individual behavioral parameters, with “D” indicating dark-phase measurements and “L” indicating light-phase measurements. e-g) Rose plots of top genes significantly different by both criteria and shared between sexes, shown for males. Plots are organized clockwise from “more awake” to“more asleep” and summarized across the full 24 h monitoring period. Red bars indicate increased behavior in MiMIC+/- flies, while blue bars indicate decreased behavior.

Genetic screening in Drosophila identifies modulators of social behavior.

a,b) Log2fold change heatmaps and significance plots for all screened genes are shown for the four social spacing parameters in males (a) and females (b). SGPVs were calculated for the comparisons (i) genotype vs wild-type and (ii) genotype vs all other mutant lines. Genes that met hit criteria (SGPV = 0) for the wild-type comparison alone are shown in gray, whereas genes that met both criteria are shown in red. Rows are ordered by the estimated effect size for Within 4 Body Lengths and displayed in descending order (n=2-3 replicates per line with 4 flies per replicate). c,d) Radar plots of the top shared (top), male-specific (middle), and female-specific (bottom) significant genes, showing t-statistics from the mutant vs wild-type contrasts in males (c) and females (d). The inner polygon represents the wild-type reference (t = 0), with axes extending outward indicating higher values relative to wild-type and axes plotting inward indicating lower values. Male and female plots share a common axis scale to allow direct comparison across genotypes.

Behavioral characterization of domMiMIC+/- flies.

a) Schematic of the dominoMIMIC+/- insertion. b) qPCR analysis of dom expression in wild-type and domMiMIC+/- flies (n=5; 10 heads per replicate; Kruskal-Wallis test). c,d) Radar plots of social spacing assay results for males (c) and females (d), showing t-statistics from the dom vs wild-type contrasts (n=16; 20 flies per replicate). The inner polygon represents the wild-type reference (t = 0), with axes extending outward indicating higher values relative to wild-type and axes plotting inward indicating lower values. Male and female plots share a common axis scale to allow for direct comparison. e) Climbing assay performance in males and females (n=16; Two-way ANOVA with Šídák multiple comparisons). f,g) PCA of activity assay results in males (f) and females (g), with wild-type (gray) and domMiMIC+/- (tan) flies (n=64). h,i) Individual sleep and activity parameters for males (h) and females (i). For each measure, the 24 h estimate is shown in the upper graph (open bars), with light- and dark-period estimates shown in the lower graph (patterned and filled bars, respectively). Estimates are plotted such that positive values indicate an increase and negative values indicate a decrease in domMiMIC+/- flies relative to controls. Asterisks indicate significance (p ≤ 0.01), with light and dark brown asterisks denoting significance during the light and dark periods, respectively.

domMiMIC+/- flies exhibit distinct transcriptional profiles in males and females.

a,b) Volcano plots of DE genes in males (a; 140 downregulated and 104 upregulated), and females (b; 147 downregulated and 413 upregulated) (n=5 replicates of 5 pooled brains; padj ≤ 0.01 and |log2fold change| ≥ 0.5). Expression of dom is significantly reduced in both sexes. c,d) GSEA of all DE genes in males (c) and females (d). e) Scatterplot of overlapping significant DE genes in males and females (tan), plotted by log2fold change (overlap significance, Fisher’s Exact Test <0.0001). f) Venn diagrams of downregulated (gray) and upregulated (tan) genes in males (left) and females (right). Numbers above circles indicate the total number of DE genes per condition, while numbers within circles denote unique or shared genes across sexes. ORA of shared genes is shown in the adjacent bar plots (padj ≤ 0.05).

Partial loss of domino alters splicing regulation and H2A.V occupancy.

a,b) Alternative splicing events detected in male (a) and female (b) adult domMiMIC+/- brains (n=5 replicates of 5 pooled brains), showing total valid events detected by rMATS and significant events (FDR ≤ 0.05 and |IncLevelDifference| ≥ 0.1). c,d) ORA of significant splicing events in males (c) and females (d). e-g) H2A.V ChIP enrichment profiles for control (gray) and dom knockdown (pink) embryos. Plots show genes with significant differential splicing identified by rMATs(46), divided into equally sized groups representing the upper (e) and lower (f) halves of the ranked gene set. Two matched groups of non-significant genes, ranked immediately below the splicing significance threshold, are shown for comparison (g), top half and (h), bottom half. Pink shading marks TSS, blue shading indicates gene bodies, and tan shading marks TES. i-l) Quantification of the mean difference in H2A.V enrichment between control and dom knockdown embryos (Two-way ANOVA mixed-effects analysis).

Partial loss of domino alters synaptic organization in domMiMIC+/- brains.

a) Immunofluorescence images of antennal lobes stained with Hoechst (blue) and nc82 (magenta), with merged images shown for both sexes. Scale bar 20µm. Brightness and contrast specifications are consistent across images. b) Fluorescence intensity was measured as mean ROI antennal lobe - mean ROI foramen, with quantification presented as a percentage of the average wild-type intensity (n=24-32; Two-way ANOVA with Šídák multiple comparisons).

Genetic effects on sleep and activity phenotypes.

a,b) Forest plots for all 68 viable screened genes in males (a) and females (b) (n=2-3 replicates per line with 4 flies per replicate). For the top 6 PCs, 95% confidence intervals were computed for two comparisons: (i) each mutant genotype versus wild-type and (ii) each genotype versus the pooled mean of all other mutant lines. SGPVs were calculated for each comparison. Genes are ranked within each PC by estimated effect size, with confidence intervals shown. Genes meeting both hit criteria (SGPV = 0) are shown in red, those meeting only the wild-type comparison are shown in gray, and inconclusive genes are shown in beige. c-e) Rose plots of top genes significantly different by both criteria and shared between sexes, shown for females. Plots are organized clockwise from “more awake” to“more asleep” and summarized across the full 24 h monitoring period. Red bars indicate increased behavior in MiMIC+/- flies, while blue bars indicate decreased behavior.

Genetic effects on social behavior phenotypes.

a,b) Forest plots for all 68 viable screened genes in males (a) and females (b) (n=2-3 replicates per line with 4 flies per replicate). SGPVs were calculated for the comparisons (i) genotype vs wild-type and (ii) genotype vs all other mutant lines. Genes are ranked within each social spacing parameter by estimated effect size, with confidence intervals shown. Genes meeting both hit criteria (SGPV = 0) are shown in red, those meeting only the wild-type comparison are shown in gray, and inconclusive genes are shown in beige.

Behavioral characterization of domTrojan+/- flies.

a) Schematic of the dominoTrojan+/- insertion. b) qPCR analysis of dom expression in wild-type and domTrojan+/- flies. (n=5; 10 heads per replicate; Kruskal-Wallis test). c,d) Radar plots of social spacing assay results for males (c) and females (d), showing t-statistics from the dom vs wild-type contrasts (n=16; 20 flies per replicate). The inner polygon represents the wild-type reference (t = 0), with axes extending outward indicating higher values relative to wild-type and axes plotting inward indicating lower values. Male and female plots share a common axis scale to allow for direct comparison. e) Climbing assay performance in males and females (n=16; Two-way ANOVA with Šídák multiple comparisons). f,g) PCA of activity assay results in males (f) and females (g), with wild-type (gray) and domTrojan+/- (teal) flies (n=64). h,i) Individual sleep and activity parameters for males (h) and females (i). For each measure, the 24 h estimate is shown in the upper graph (open bars), with light- and dark-period estimates shown in the lower graph (patterned and filled bars, respectively). Estimates are plotted such that positive values indicate an increase and negative values indicate a decrease in domMiMIC+/- flies relative to controls. Asterisks indicate significance (p ≤ 0.01), with light and dark blue asterisks denoting significance during the light and dark periods, respectively. j-m) Total pupal counts from dom RNAi crosses targeting (j) tubulin-Gal4 (whole-body), (k) repo-Gal4 (glia), (l) elav-Gal4 (neurons) and (m) MB-Gal4 (mushroom body). Light gray open circles represent pupae that were wild-type due to the presence of a balancer chromosome. n-q) Percentage of total pupae that successfully eclosed to adults for the same RNAi crosses shown in j-m (Mann-Whitney test).

domTrojan+/- flies exhibit distinct transcriptional profiles in males and females.

a,b) Volcano plots of DE genes in males (a; 113 downregulated and 141 upregulated), and females (b; 161 downregulated and 166 upregulated) (n=5 replicates of 5 pooled brains; padj ≤ 0.01 and |log2fold change| ≥ 0.5). Expression of dom is significantly reduced in both sexes. c,d) GSEA of all DE genes in males (c) and females (d). e) Scatterplot of overlapping significant DE genes in males and females (teal), plotted by log2fold change (overlap significance, Fisher’s Exact Test <0.0001). f) Venn diagrams of downregulated (gray) and upregulated (teal) genes in males (left) and females (right). Numbers above circles indicate the total number of DE genes per condition, while numbers within circles denote unique or shared genes across sexes.

Partial loss of domino alters RNA splicing and synaptic organization in domTrojan+/- brains.

a,b)) Alternative splicing events detected in male (a) and female (b) adult domTrojan+/-brains (n=5 replicates of 5 pooled brains), showing total valid events detected by rMATS and significant events (FDR ≤ 0.05 and |IncLevelDifference| ≥ 0.1). c,d) ORA of significant splicing events in males (c) and females (d). e) Immunofluorescence images of antennal lobes stained with Hoechst (blue) and nc82 (magenta), with merged images shown for both sexes. Brightness and contrast specifications are consistent across images. Scale bar 20µm. f) Fluorescence intensity was measured as mean ROI antennal lobe - mean ROI foramen, with quantification presented as a percentage of the average wild-type intensity (n=23-34; Two-way ANOVA with Šídák multiple comparisons).