MOESM1 of A meta-analysis of two high-risk prospective cohort studies reveals autism-specific transcriptional changes to chromatin, autoimmune, and environmental response genes in umbilical cord blood

Additional file 1: Figure S1. Overview of study design. Figure S2. Surrogate variable analysis in MARBLES subjects. Figure S3. Surrogate variable analysis in EARLI subjects for the ASD versus TD comparison. Figure S4. Surrogate variable analysis in EARLI subjects for the Non-TD versus TD comparison. Figure S5. Identification of ASD-associated differentially-expressed genes in cord blood within each study. Figure S6. Identification of Non-TD-associated differentially-expressed genes in cord blood within each study. Figure S7. Correlations between ASD and Non-TD expression differences in MARBLES and EARLI subjects. Figure S8. Expression level distribution of meta-analysis ASD versus TD differential probes is similar to non-differential probes. Figure S9. Cord blood differentially-expressed genes are not enriched for ASD-associated gene sets. Figure S10. Cord blood differentially-expressed genes are depleted for blood cell-specific genes. Figure S11. Expression level distribution of meta-analysis Non-TD versus TD differential probes is similar to non-differential probes. Figure S12. Consensus coexpression modules identified in MARBLES and EARLI. Figure S13. Consensus module eigengene networks are preserved between MARBLES and EARLI subjects. Figure S14. Consensus modules are correlated with diagnosis and demographic factors in MARBLES subjects. Figure S15. Consensus modules are correlated with demographic factors in EARLI subjects. Figure S16. Consensus modules are correlated with diagnosis and demographic factors in meta-analysis. Figure S17. Skyblue1 module is specifically expressed in males and is enriched for genes upregulated in ASD. Figure S18. Consensus modules are strongly correlated with cell type proportions in meta-analysis. Figure S19. The B cell-associated grey60 module is upregulated during cesarean delivery and is downregulated in ASD. Figure S20. Anticipated power curve calculated by ssize.fdr R package with projected proportions of non-differentially expressed genes as 0.8, 0.85, 0.9, 0.95 and 0.98.