Figure 1
Copy number variant processing workflow. (A) The copy number variant (CNV) calling pipeline applied to Infinium Global Screening Array genotyping data from individuals with ET (n = 1,853) and controls (n = 10,336). (B) Quality control steps applied to the called CNVs.
Figure 2
Global burden analyses of rare copy number variants encompassing protein coding regions in individuals with ET (n = 1,204) compared to controls (n = 9,549). Forest plot depicting logistic regression results from tests of ET status with number of genes affected by copy number variants (CNVs), number of CNVs, and CNV size (measured in units of 100kbs) between ET patients and controls are shown. Odds ratios and 95% confidence intervals for each burden metric regression analyses are shown in red for deletions and in blue for duplications. P-values are also displayed. Bonferroni adjusted p-value cut-off: 0.05/6.
Figure 3
Global burden analyses of rare copy number variants encompassing protein coding regions segregated by size. Forest plot depicting logistic regression results from tests of ET status with number of genes affected by copy number variants (CNVs) and number of CNVs between ET patients and controls are shown for different CNV size cut-offs. CNVs are divided in 4 subsets based on size, these are CNVs: < 100kbs, 100kbs – 500kbs, 500kbs – 1MB, and > 1MBs. Odds ratios and 95% confidence intervals for each burden metric regression analyses are shown in red for deletions and in blue for duplications. P-values are also displayed. Bonferroni adjusted p-value cut-off for deletions: 0.05/8. Bonferroni adjusted p-value cut-off for duplications: 0.05/8.
Figure 4
Gene-set enrichment analyses of rare copy number variants encompassing protein coding regions. Forest plot depicting logistic regression results from tests of ET status with copy number variant (CNV) counts across gene-sets are shown. Odds ratios and 95% confidence intervals for each burden metric regression analyses are shown in red for deletions and in blue for duplications. No CNVs belonging to cases nor controls possessed duplications that affected any gene belonging to the rare CNV (rCNV) triplosensitive or haploinsufficient and ClinGen triplosensitive or haploinssuficent gene-sets, nor did any CNV affect any gene in the neurodegenerative gene-set. These gene-sets are therefore not depicted in the figure. Bonferroni adjusted p-value cut-off was: 0.05/30.
Figure 5
Pathogenic prediction scoring burden analyses of rare copy number variants encompassing protein coding regions. Forest plot depicting logistic regression results from tests of ET status with average pathogenicity score burden associations of Functional Analysis Through Hidden Markov Models (FATHMM), logistic regression (LR), Variant Effect Scoring Tool (VEST3), Combined Annotation Dependent Depletion (CADD), Probability of being Loss-of-function Intolerant (pLI), and meta-voting prediction (MVP) scores are shown. Odds ratios and 95% confidence intervals for each burden metric regression analyses are shown in red for deletions and in blue for duplications. P-values are also displayed. Bonferroni adjusted p-value cut-off for deletions and duplications: 0.05/12.