Abstract
Background:
Genetic factors contribute to the development of anxiety disorders, yet few risk genes have been previously identified. One genomic approach that has achieved success in identifying risk genes in related childhood neuropsychiatric conditions is investigations of de novo variants, which has yet to be leveraged in childhood anxiety disorders.
Methods:
We performed whole-exome DNA sequencing in 76 parent-child trios (68 trios after quality control) recruited from a childhood anxiety disorder clinic and compared rates of rare and ultra-rare de novo variants with 790 previously sequenced control trios (783 trios after quality control). We then explored overlap with risk genes for other neuropsychiatric conditions and enrichment in biologic pathways.
Results:
Rare and ultra-rare de novo likely gene disrupting and predicted damaging missense genetic variants are enriched in anxiety disorder probands compared with controls (rare variant rate ratio 1.97, 95% CI 1.11–3.34, p=0.03; ultra-rare variant rate ratio 2.59, 95% CI 1.35–4.70, p=0.008). These de novo damaging variants occur in individuals with a variety of childhood anxiety disorders and impact genes that have been associated with other neuropsychiatric conditions. Exploratory network analyses reveal enrichment of deleterious variants in canonical biological pathways.
Conclusions:
These findings provide a path for identifying risk genes and promising biologic pathways in childhood anxiety disorders by de novo genetic variant detection. Our results suggest the discovery potential of applying this approach in larger anxiety disorder cohorts to advance our understanding of the underlying biology of these common and debilitating conditions.
Keywords: anxiety disorders, childhood, genomics, exome sequencing, de novo, rare variants, biologic pathways
1. INTRODUCTION
Anxiety disorders comprise the most common class of psychiatric conditions in childhood (Kessler, Petukhova, Sampson, Zaslavsky, & Wittchen, 2012; Merikangas et al., 2010), and therefore may represent some of the earliest forms of psychopathology for etiologic investigation (Beesdo, Knappe, & Pine, 2009). Family and twin studies have long demonstrated the important role of genetic factors in the development of anxiety disorders (John M. Hettema, 2001; Lawrence, Murayama, & Creswell, 2019; Scaini, Belotti, & Ogliari, 2014; Scaini, Ogliari, Eley, Zavos, & Battaglia, 2012; Smoller, 2016). These studies suggest genetic overlap among different anxiety disorders, with evidence that the genetic components in common may be greater than those unique to a specific anxiety disorder (Tambs et al., 2009). Furthermore, the contribution of genetic factors to anxiety disorders may change over the course of development, with a larger genetic contribution for anxiety disorders that emerge in childhood (Kendler, Gardner, Annas, & Lichtenstein, 2008; Scaini et al., 2014; Waszczuk, Zavos, Gregory, & Eley, 2014), highlighting the discovery potential of genomic investigations focused on childhood anxiety disorders.
Although the important role of genetic factors in the development of anxiety disorders is well established, progress in identifying specific reproducible risk genes has been slow. Recent large-scale genome-wide association studies have identified a few significant genetic variants for anxiety disorders (Levey et al., 2020; Meier et al., 2019; Otowa et al., 2016; Purves et al., 2019). However, effect sizes of these variants and their explained proportion of the variability in anxiety disorders are small. Research focused on rare de novo (spontaneous) genetic variants with larger effect sizes has enabled the discovery of multiple high-confidence risk genes in several related childhood-onset neuropsychiatric conditions, including autism spectrum disorder, Tourette’s disorder, and obsessive-compulsive disorder (Cappi et al., 2020; Satterstrom et al., 2020; S. Wang et al., 2018). But this powerful approach of examining de novo genetic variants has not yet been leveraged in the field of anxiety disorders.
In this study, we conducted the first whole-exome DNA sequencing study focused on parent-child trios impacted by childhood anxiety disorders in order to identify rare and ultra-rare de novo genetic variants. Sequencing was performed in 76 parent-child trios (228 individuals total) recruited from the Anxiety and Mood Disorders Program at the Yale Child Study Center and compared to 790 previously sequenced control trios without anxiety disorders. Based on studies of other childhood neuropsychiatric conditions, we hypothesized that likely gene disrupting (LGD) variants (including single nucleotide variants introducing premature stop codons, frameshift indels, and single nucleotide splice site variants) and missense variants that were predicted to be damaging (Mis-D) would be increased in cases versus controls. Furthermore, using the list of genes with de novo damaging variants in the anxiety disorder cases, we assessed for overlap with previously identified risk genes in other neuropsychiatric conditions. We also conducted exploratory analyses to identify biological pathways (such as molecular functions or biologic processes) that are enriched in this list of genes with de novo damaging variants more than expected by chance. The results from this study provide novel insight into the role of gene-damaging de novo mutations in childhood anxiety disorders. This is an important step towards improving our etiologic understanding of childhood anxiety disorders and developing future treatments of these common and impairing conditions.
2. METHODS
2.1. Procedures
This study was approved by the Institutional Review Board at the Yale School of Medicine and conforms to standards of the US Federal Policy for the Protection of Human Subjects. Following initial telephone screening, children and parents were invited for in-person evaluations at the Anxiety and Mood Disorders Program at the Yale Child Study Center between November 2016 and November 2019. Children were excluded at the telephone screening stage if they had autism spectrum disorder or intellectual disability. Informed consent/assent was obtained from all parents and children enrolled in this study. The Anxiety Disorders Interview Schedule for Children (Child and Parent versions; ADIS-C/P), administered by trained and experienced interviewers, was used to assess and diagnose the children’s anxiety disorders (Silverman, Saaverdram, & Pina, 2001). Sampling was not restricted to a specific childhood anxiety disorder given the high rates of comorbidity and evidence supporting shared genetic etiology (Franco, Saavedra, & Silverman, 2007; Tambs et al., 2009).
2.2. Participants
Whole-exome DNA sequencing was performed in 76 parent-child trios recruited from the Anxiety and Mood Disorders Program at Yale. After employing several quality control methods (described below), our sample size was 68 anxiety disorder parent-child trios. The probands from these 68 parent-child trios had an average age of 9 years at participation, were 57% male, and the majority identified as white (85%) (Table 1). Many of these children met criteria for multiple anxiety disorders as is typical of clinical samples of childhood anxiety disorders. The most common diagnoses were generalized anxiety disorder (74%), social phobia (51%), separation anxiety disorder (46%), and specific phobia (28%). The most common comorbid psychiatric diagnoses were attention-deficit/hyperactivity disorder (22%), oppositional defiant disorder (16%), dysthymia (12%), and obsessive-compulsive disorder (10%). One proband did not meet full criteria for an anxiety disorder diagnosis based on the ADIS-C/P and was excluded from the analysis. For the controls, whole-exome DNA sequencing data from 790 unaffected parent-child trios were selected from the Simons Simplex Collection from the National Institutes of Health Data Archive (https://ndar.nih.gov/edit_collection.html?id=2042). These control parent-child trios were selected to be in the normal range for the anxiety problems subscale from the Child Behavioral Checklist (CBCL) or Adult Behavioral Checklist (ABCL) (t-score<64.5), which has been shown to predict anxiety disorder diagnoses (Knepley, Kendall, & Carper, 2019). All control subjects also had no evidence of autism spectrum disorder (Fischbach & Lord, 2010).
Table 1:
Characteristics of 68 children with anxiety disorders included in de novo variant analysis
| Characteristics | Anxiety probands (n=68) |
|---|---|
| Age at interview, mean ± sd | 9.2 ± 2.7 |
| Gender | |
| female, n (%) | 29 (43%) |
| male, n (%) | 39 (57%) |
| Race | |
| White, n (%) | 58 (85%) |
| Asian, n (%) | 2 (3%) |
| Black or African American, n (%) | 1 (1%) |
| Multi-racial, n (%) | 6 (9%) |
| Ethnicity | |
| Hispanic or Latino, n (%) | 10 (15%) |
| Childhood anxiety disorder diagnoses, n (%) | |
| Generalized anxiety disorder | 50 (74%) |
| Social phobia | 35 (51%) |
| Separation anxiety disorder | 31 (46%) |
| Specific phobia | 19 (28%) |
| Agoraphobia | 1 (1%) |
| Panic disorder | 1 (1%) |
| Selective mutism | 1 (1%) |
| Other co-morbid diagnoses, n (%) | |
| Attention-deficit/hyperactivity disorder | 15 (22%) |
| Opposition defiant disorder | 11 (16%) |
| Dysthymia | 8 (12%) |
| Obsessive-compulsive disorder | 7 (10%) |
| Sleep terror disorder | 2 (3%) |
| Major depressive disorder | 1 (1%) |
| Post-traumatic stress disorder | 1 (1%) |
2.3. Whole-exome DNA sequencing
Exome capture and whole-exome sequencing of saliva-derived DNA was performed at the Yale Center for Genome Analysis (YCGA) for the 76 anxiety disorder parent-child trios (228 samples). The NimbleGen SeqCap MedExome capture library (Roche NimbleGen, Madison, WI) was used for the anxiety samples. A combination of the Illumina HiSeq 4000 (74-bp paired-end reads) and NovaSeq platforms (Illumina, San Diego, CA) were used for sequencing the anxiety samples (HiSeq 4000 for 69 samples and NovaSeq for 159 samples), which were multiplexed in each capture reaction and sequencing lane. This data was pooled with 790 control parent-child trios without anxiety disorders, already sequenced as part of the Simons Simplex Collection, for joint variant calling. These blood-derived control samples were sequenced on the Illumnia HiSeq 2000 sequencing platform after capture with the NimbleGen SeqCap EzExomeV2 library.
2.4. Sequence alignment and variant identification
Alignment and variant calling of the sequencing reads followed the Genome Analysis Toolkit (GATK) best practice guidelines (McKenna et al., 2010) as recently described by our group (Cappi et al., 2020). To minimize downstream effects of differential coverage between the two different capture platforms used between cases and controls, a target bed file was created by taking the intersection of the EZExomeV2 and MedExome target regions. Case and control samples were called jointly using GATK’s GenotypeGVCFs tool and variant score recalibration was applied to these called variants. Only passing variants were used in downstream analyses. Variants were annotated using the RefSeq hg19 gene definitions and external databases using ANNOVAR (K. Wang, Li, & Hakonarson, 2010).
2.5. Quality control for de novo variant identification
Relatedness statistics were calculated based on the method of Manichaikul et al. (2010), and trios were omitted if unexpected family relationships were identified. Trios were also excluded if the children were observed to have an outlier number (>16) of de novo variants. PLINK/SEQ istats (https://psychgen.u.hpc.mssm.edu/plinkseq/stats.shtml) was used to generate quality control metrics for both cases and controls, which were used to generate principal components to identify outliers that were removed from the analysis (see Figure S1 and Table S1 for details). After these quality control steps were employed, we analyzed 68 anxiety disorder parent-child trios and 783 control parent-child trios. We then used stringent thresholds to identify de novo mutations as previously described (Cappi et al., 2020), including that the child was heterozygous for the variant with alternate allele frequency between 0.3 and 0.7 in the child and <0.05 in the parents, sequencing depth ≥20 in all family members at the variant position, alternate allele depth ≥5, observed allele frequency <0.01 among all cases and controls, and mapping quality ≥30. In our analysis, we examined both rare de novo variants that were present with a frequency of <0.001 (0.1%) in the Genome Aggregation Database (gnomAD) (Karczewski et al., 2020) as well as ultra-rare de novo variants that were not present in gnomAD. The v2.1.1 gnomAD data set contains 125,748 harmonized exome sequences and 15,708 whole-genome sequences from unrelated individuals.
2.6. Mutation rate analysis
De novo mutations per base pair were calculated in cases and controls. To ensure accurate rate calculations, we determined the denominator of “callable” base pairs per family using the GATK DepthOfCoverage tool, requiring that bases were covered at >=20x in all family members, base quality >=20, and map quality>=30 to match GATK thresholds. We focused on variants in the callable exome to further minimize any potential calling bias between the cases and controls. The resulting rates were divided by two to give haploid rates, and confidence intervals were calculated (pois.conf.int, pois.exact function from the epitools v0.5–9 in R). We compared de novo mutation rates in cases versus controls using a one-tailed rate ratio test (rateratio.test package in R). Based on studies of other childhood onset neuropsychiatric conditions (Cappi et al., 2020; Halvorsen et al., 2021; Satterstrom et al., 2020; S. Wang et al., 2018), we hypothesized that rare de novo likely gene disrupting (LGD) variants (including single nucleotide variants introducing premature stop codons, frameshift indels, and single nucleotide splice site variants) and missense variants that were predicted to be damaging (Mis-D) would be enriched in cases vs controls. We used the integrated “missense badness, PolyPhen-2, constraint” (MPC) (Samocha et al., 2017) score of >2 to identify Mis-D variants as done in other recent whole-exome sequencing studies (Feliciano et al., 2019; Satterstrom et al., 2020). The combined group of LGD and Mis-D variants was considered de novo damaging variants. All de novo damaging variants in the anxiety disorder parent-child trios were confirmed by in silico visualization.
2.7. Gene set overlap
The Gene4Denovo database (Zhao et al., 2020) (http://www.genemed.tech/gene4denovo/home) was used to assess whether genes with de novo damaging variants (either LGD or Mis-D) in the anxiety disorder parent-child trios overlapped with genes previously associated with other disorders in other sequencing studies. Gene4Denovo integrates de novo mutations from 58,011 individuals across 28 different phenotypes and prioritizes a list of candidate genes. These phenotypes include several neuropsychiatric conditions, including autism spectrum disorder, schizophrenia, intellectual disability, epileptic encephalopathies, undiagnosed developmental disorders, Tourette’s disorder, bipolar disorder, obsessive-compulsive disorder, and others. Specifically, we assessed the overlap between the Gene4Denovo candidate gene list (Release version 1.1, updated 09/09/2020) with a false discovery rate (FDR)<0.05 and our list of genes with damaging variants (13 genes total).
2.8. Exploratory Pathway Analyses
We used ConsensusPathDB (Herwig, Hardt, Lienhard, & Kamburov, 2016) (http://cpdb.molgen.mpg.de/, Release 35, 05.06.2021) to assess if the same list of anxiety disorder genes harboring de novo damaging variants used in the gene set overlap analysis were enriched for biological pathways. This network interaction tool integrates data from 31 public resources and literature curation. The following default settings were used for the gene set over-representation analysis with gene identifier type of Ensembl; Pathway based sets: pathways as defined by pathway databases, select all resources, minimum overlap with input list=2, p-value cutoff=0.01; Gene ontology categories: gene ontology level 2 category, select all categories, p-value cutoff=0.01.
3. RESULTS
3.1. De novo damaging DNA sequence variants are associated with childhood anxiety disorders
Whole-exome DNA sequencing was performed in 76 parent-child trios recruited from the Anxiety and Mood Disorders Program at Yale. These data were pooled with 790 control parent-child trios without clinically significant anxiety already sequenced as part of the Simons Simplex Collection for joint variant calling. After employing quality control methods, our sample size for burden analysis was 68 anxiety disorder parent-child trios and 783 control trios. Based on studies of other childhood neuropsychiatric conditions (Cappi et al., 2020; Halvorsen et al., 2021; Satterstrom et al., 2020; S. Wang et al., 2018), we expected to find an enrichment of rare de novo damaging variants, including LGD variants (premature stop codons, frameshift mutations, and splice site variants) and missense variants predicted to be damaging by an MPC score>2 (Mis-D).
In our burden analysis, rare and ultra-rare de novo damaging variants (LGD+Mis-D) were significantly enriched in cases compared to controls. For rare de novo damaging variants (gnomAD allele frequency<0.001), the rate ratio for cases vs controls was 1.97 (95% CI 1.11–3.34, p=0.03) (Figure 1A, Table 2, Table S2). For ultra-rare de novo damaging variants (not present in gnomAD), this rate ratio was 2.59 (95% CI 1.35–4.70, p=0.008) (Figure 1B, Table 3). When we examined classes of de novo damaging variants separately, there was a trend towards an increased rate of rare Mis-D variants in cases vs controls (rate ratio 2.36, 95% CI 0.84–5.71, p=0.09) and rare LGD variants in cases vs controls (rate ratio 1.79, 95% CI 0.84–3.50, p=0.10). For the ultra-rare de novo variants, the Mis-D variants were significantly enriched in cases vs controls (rate ratio 3.58, 95% CI 1.04–10.52, p=.04) and there was a trend towards an increased rate for LGD variants (rate ratio 2.24, 95% CI 0.97–4.68, p=.06). As expected, we did not detect a difference in mutation rates across all de novo variants (including those not expected to be damaging) or in synonymous de novo variants between cases and controls (Figure 1, Tables 2 and 3).
Figure 1: Rare and ultra-rare de novo damaging variants are enriched in anxiety parent-child trios (n=68) compared with controls (n=783).
Bar chart comparing rates of (A) rare and (B) ultra-rare de novo mutation types between anxiety disorder cases (red) and controls (blue). Rare variants have an allele frequency (AF) of <0.001 (0.1%) in the Genome Aggregation Database (gnomAD) and ultra-rare de novo variants are not present in gnomAD. The mutation rate per base pair (bp) considers only those “callable” loci in each family and cohort that meet required sequencing depth and quality scores. Mutation rates were compared using a 1-tailed R rateratio.test with a p<0.05 considered significant. Mis-D: missense variants predicted to be damaging with MPC “missense badness, PolyPhen-2, constraint” score >2. LGD: likely gene disrupting, including frameshift indels and stop-gain variants. Synonymous: a single nucleotide variant that does not alter the amino acid sequence.
Table 2:
Distribution of classes of rare de novo variants (gnomAD AF<0.001) in anxiety disorder cases and controls
| De novo variant classa | Variant counts | Mutation Rate (×10−8) per BP (95% CI)b | Estimated Coding Variants per Individual (95% CI)c | Rate Ratio (95% CI)d | p value | |||
|---|---|---|---|---|---|---|---|---|
| Anxiety Cases (n=68) | Controls (n=783) | Anxiety Cases (n=68) | Controls (n=783) | Anxiety Cases (n=68) | Controls (n=783) | |||
| Synonymous SNVs | 14 | 171 | 0.34 (0.19–0.57) | 0.46 (0.40–0.54) | 0.23 (0.13–0.39) | 0.31 (0.27–0.36) | 0.73 (0.44–1.17) | 0.90 |
| All Missense (Mis)e | 48 | 431 | 1.16 (0.86–1.54) | 1.17 (1.06–1.28) | 0.79 (0.58–1.04) | 0.79 (0.72–0.87) | 1.00 (0.76–1.29) | 0.53 |
| Missense with MPC 0–1 (Mis-B)f | 29 | 302 | 0.7 (0.47–1.01) | 0.82 (0.73–0.91) | 0.48 (0.32–0.68) | 0.55 (0.49–0.62) | 0.86 (0.61–1.19) | 0.81 |
| Missense with MPC 1–2 (Mis-P)g | 8 | 81 | 0.19 (0.08–0.38) | 0.22 (0.17–0.27) | 0.13 (0.06–0.26) | 0.15 (0.12–0.18) | 0.88 (0.43–1.66) | 0.68 |
| Missense with MPC >2 (Mis-D)h | 5 | 19 | 0.12 (0.04–0.28) | 0.05 (0.03–0.08) | 0.08 (0.03–0.19) | 0.03 (0.02–0.05) | 2.36 (0.84–5.71) | 0.09 |
| All LGDi | 8 | 40 | 0.19 (0.08–0.38) | 0.11 (0.08–0.15) | 0.13 (0.06–0.26) | 0.07 (0.05–0.10) | 1.79 (0.84–3.50) | 0.10 |
| LGD frameshift indels | 5 | 21 | 0.12 (0.04–0.28) | 0.06 (0.04–0.09) | 0.08 (0.03–0.19) | 0.04 (0.02–0.06) | 2.13 (0.77–5.10) | 0.11 |
| LGD SNVs | 3 | 19 | 0.07 (0.01–0.21) | 0.05 (0.03–0.08) | 0.05 (0.01–0.14) | 0.03 (0.02–0.05) | 1.41 (0.36–4.14) | 0.38 |
| Nonframeshift indels | 2 | 9 | 0.05 (0.01–0.18) | 0.02 (0.01–0.05) | 0.03 (0–0.12) | 0.02 (0.01–0.03) | 1.99 (0.31–7.95) | 0.30 |
| Damaging (LGD + MIS-D) |
13 | 59 | 0.31 (0.17–0.54) | 0.16 (0.12–0.21) | 0.21 (0.11–0.36) | 0.11 (0.08–0.14) | 1.97 (1.11–3.34) | 0.03 |
| All variantsj | 72 | 656 | 1.74 (1.36–2.20) | 1.77 (1.64–1.92) | 1.18 (0.92–1.49) | 1.20 (1.11–1.30) | 0.98 (0.79–1.21) | 0.57 |
gnomAD, the Genome Aggregation Database; AF, allele frequency; CI, confidence interval; SNVs, single nucleotide variant’s LGD, likely gene disrupting; indel, insertion-deletion variant; Mis-B, benign missense; Mis-P, possibly damaging missense; Mis-D, damaging missense;
Variants were annotated with ANNOVAR using RefSeq hg19 definitions.
De novo mutation rates were calculated as the number of variants divided by the number of haploid “callable” bae pairs.
The estimated number of de novo mutations per individual was calculated by multiplying by the size of the RefSeq hg19 coding exome (33,828,798 bp).
Rates were compared using a 1-sided rate ratio test with p value <0.05 considered significant.
All missense variants contain 6 missense variants in anxiety disorder cases and 30 in control trios that were not annotated by the “missense badness, PolyPhen-2, constraint” scores (MPC).
Missense variants with an MPC score of 0–1.
Missense variants with an MPC score of 1–2.
Missense variants with an MPC score of >2.
Includes variants altering a stop codon and frameshift indels (no changes to canonical splice sites were identified in this analysis).
All annotated coding variants, which includes 5 variants annotated as unknown in controls.
Table 3:
Distribution of classes of ultra-rare de novo variants (gnomAD AF=0) in anxiety disorder cases and controls
| De novo variant classa | Variant counts | Mutation Rate (×10−8) per BP (95% CI)b | Estimated Coding Variants per Individual (95% CI)c | Rate Ratio (95% CI)d | p value | |||
|---|---|---|---|---|---|---|---|---|
| Anxiety Cases (n=68) | Controls (n=783) | Anxiety Cases (n=68) | Controls (n=783) | Anxiety Cases (n=68) | Controls (n=783) | |||
| Synonymous SNVs | 10 | 85 | 0.34 (0.19–0.57) | 0.23 (0.18–0.28) | 0.23 (0.13–0.39) | 0.16 (0.12–0.19) | 1.05 (0.55–1.86) | 0.49 |
| All Missense (Mis)e | 27 | 262 | 1.16 (0.86–1.54) | 0.71 (0.63–0.80) | 0.79 (0.58–1.04) | 0.48 (0.42–0.54) | 0.92 (0.64–1.30) | 0.68 |
| Missense with MPC 0–1 (Mis-B)f | 13 | 182 | 0.70 (0.47–1.01) | 0.49 (0.42–0.57) | 0.48 (0.32–0.68) | 0.33 (0.29–0.39) | 0.64 (0.37–1.04) | 0.96 |
| Missense with MPC 1–2 (Mis-P)g | 7 | 52 | 0.19 (0.08–0.38) | 0.14 (0.11–0.18) | 0.13 (0.06–0.26) | 0.10 (0.07–0.12) | 1.21 (0.54–2.40) | 0.38 |
| Missense with MPC >2 (Mis-D)h | 4 | 10 | 0.12 (0.04–0.28) | 0.03 (0.01–0.05) | 0.08 (0.03–0.19) | 0.02 (0.01–0.03) | 3.58 (1.04–10.52) | 0.04 |
| All LGDi | 7 | 28 | 0.19 (0.08–0.38) | 0.08 (0.05–0.11) | 0.13 (0.06–0.26) | 0.05 (0.03–0.07) | 2.24 (0.97–4.68) | 0.06 |
| LGD frameshift indels | 4 | 16 | 0.12 (0.04–0.28) | 0.04 (0.02–0.07) | 0.08 (0.03–0.19) | 0.03 (0.02–0.05) | 2.24 (0.69–6.00) | 0.13 |
| LGD SNVs | 3 | 12 | 0.07 (0.01–0.21) | 0.03 (0.02–0.06) | 0.05 (0.01–0.14) | 0.02 (0.01–0.04) | 2.24 (0.54–7.03) | 0.19 |
| Nonframeshift indels | 2 | 7 | 0.05 (0.01–0.18) | 0.02 (0.01–0.04) | 0.03 (0.00–0.12) | 0.01 (0.01–0.03) | 2.56 (0.38–10.93) | 0.23 |
| Damaging (LGD + MIS-D) |
11 | 38 | 0.31 (0.17–0.54) | 0.10 (0.07–0.14) | 0.21 (0.11–0.36) | 0.07 (0.05–0.10) | 2.59 (1.35–4.70) | 0.008 |
| All variantsj | 46 | 386 | 1.74 (1.36–2.20) | 1.04 (0.94–1.15) | 1.18 (0.92–1.49) | 0.71 (0.64–0.78) | 1.07 (0.81–1.39) | 0.36 |
gnomAD, the Genome Aggregation Database; AF, allele frequency; CI, confidence interval; SNVs, single nucleotide variant’s LGD, likely gene disrupting; indel, insertion-deletion variant; Mis-B, benign missense; Mis-P, possibly damaging missense; Mis-D, damaging missense;
Variants were annotated with ANNOVAR using RefSeq hg19 definitions.
De novo mutation rates were calculated as the number of variants divided by the number of haploid “callable” bae pairs.
The estimated number of de novo mutations per individual was calculated by multiplying by the size of the RefSeq hg19 coding exome (33,828,798 bp).
Rates were compared using a 1-sided rate ratio test with p value <0.05 considered significant.
All missense variants contain 3 missense variants in anxiety disorder cases and 19 in control trios that were not annotated by the “missense badness, PolyPhen-2, constraint” scores (MPC).
Missense variants with an MPC score of 0–1.
Missense variants with an MPC score of 1–2.
Missense variants with an MPC score of >2.
Includes variants altering a stop codon and frameshift indels (no changes to canonical splice sites were identified in this analysis).
All annotated coding variants, which includes 4 variants annotated as unknown in controls.
Among 68 anxiety disorder parent-child trios that passed quality control, we identified 13 de novo damaging variants in 12 children (Table 4). Similar to the characteristics of the entire anxiety disorder sample, these children with identified damaging de novo variants had a variety of childhood anxiety disorders, including generalized anxiety disorder (9/12), separation anxiety disorder (6/12), specific phobia (5/12), and social phobia (3/12). Seven of these children met criteria for other psychiatric diagnoses with the most common being attention-deficit/hyperactivity disorder (3/12).
Table 4:
Rare de novo damaging variants identified in anxiety disorder parent-child trios
| ID | Sexa | Ageb | Anxiety disordersc | Other disorders | Gened | Protein change | Genomic coordinate (hg19) | Ref. allele | Alt. allele | Variant classe | gnomAD AFf | Risk genes for other conditionsg |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Anx22.p1 | M | 11 | GAD | OCD | KIAA1429 (VIRMA) | p.P141fs | Chr8: 95547128 | T | TG | Frameshift insertion | 0 | |
| Anx23.p1 | F | 13 | GAD, SAD | PTSD | SEMA3F | p.C372Y | Chr3: 50220879 | G | A | Missense (MPC 2.39) | 0 | |
| Anx28.p1 | M | 9 | GAD | ADHD | ATRX | p.Y1770X | ChrX: 76874298 | A | C | Stopgain SNV | 0 | |
| Anx34.p1 | F | 12 | Specific phobia | CACNA1A | p.N1379fs | Chr19: 13372382 | T | TG | Frameshift insertion | 0 | developmental disorders; epileptic encephalopathies | |
| ITPKB | p.L752P | Chr1: 226829818 | A | G | Missense (MPC 2.24) | 0 | ||||||
| Anx36.p1 | M | 7 | Specific phobia, SAD, GAD | ADHD, ODD | TET1 | p.Q453X | Chr10: 70333452 | C | T | Stopgain SNV | 0 | |
| Anx37.p1 | F | 7 | SOP | PCLO | p.R4461X | Chr7: 82538249 | G | A | Stopgain SNV | 0 | ||
| Anx41.p1 | F | 9 | Specific phobia, GAD, SAD | ADHD | FZD4 | p.D428fs | Chr11: 86662512 | TTGTC | T | Frameshift deletion | 7.95×10−6 | |
| Anx50.p1 | F | 6 | SAD, specific phobia, GAD | CANX | p.R570fs | Chr5: 179153744 | GA | G | Frameshift deletion | 0 | ||
| Anx51.p1 | M | 8 | SAD, specific phobia | MED7 | p.Q200fs | Chr5: 156565843 | CT | C | Frameshift deletion | 0 | ||
| Anx52.p1 | F | 14 | SOP, GAD | Dysthymia | PPP2R1A | p.R260H | Chr19: 52716335 | G | A | Missense (MPC 2.92) | 0 | developmental disorder |
| Anx59.p1 | F | 15 | GAD, SOP | Dysthymia | BPTF | p.E2895K | Chr17: 65972031 | G | A | Missense (MPC 2.47) | 0 | |
| Anx71.p1 | M | 7 | GAD, SAD | MAP3K10 | p.L322P | Chr19: 40710493 | T | C | Missense (MPC 2.38) | 3.98×10−6 |
Ref. allele, reference allele; Alt. allele, alternative allele; gnomAD, the Genome Aggregation Database; AF, allele frequency; SAD, separation anxiety disorder; GAD, generalized anxiety disorder; SOP, social phobia; OCD, obsessive-compulsive disorder; PTSD, post-traumatic stress disorder; ADHD, attention-deficit/hyperactivity disorder; ODD, oppositional defiant disorder; MPC, “missense badness, PolyPhen-2, constraint” scores; SNV, single nucleotide variant.
Sex reported as M (male) and F (female).
Age in years at assessment.
The Anxiety Disorder Interview Schedule (ADIS-C/P) was used to determine anxiety disorder diagnosis.
Genetic variants were annotated with ANNOVAR using RefGene hg19 definitions.
Damaging de novo variants were defined as likely gene disrupting variants (frameshift insertions, frameshift deletions, stopcodon change) or missense variants predicted to be damaging by a MPC “missense badness, PolyPhen-2, constraint” score >2. For missense variants, the MPC score of the variant is listed in parentheses.
The allele frequency of each variant in gnomAD (the Genome Aggregation Database).
The Gene4Denovo database (Zhao et al., 2020) (http://www.genemed.tech/gene4denovo/home) was used to assess if genes with de novo damaging variants in the anxiety disorder parent-child trios overlapped with genes previously associated with other disorders in other sequencing studies based on a False Discovery Rate (FDR)<0.05.
3.2. Genes with de novo damaging variants overlap with risk genes for other neuropsychiatric conditions
Using our list of genes with de novo damaging variants in the 68 anxiety disorder parent-child trios (Table 4), we queried the Gene4Denovo database (Zhao et al., 2020) (http://www.genemed.tech/gene4denovo/home) to identify any overlap with previously identified high-confidence risk genes for other conditions. We found that CACNA1A was a previously identified risk gene for developmental disorders (FDR=0 with 4 LGD and 47 missense damaging de novo variants previously identified) and epileptic encephalopathies (FDR=4.23×10−6 with 4 damaging missense de novo variants previously identified). We also found that the gene PPP2R1A was a risk gene for developmental disorders (FDR=5.35×10−17 with 16 damaging missense de novo variants previously identified). This highlights an overlap between genes with damaging de novo variants in children with anxiety disorders and risk genes for other neurodevelopmental conditions. Of note, these children with de novo damaging variants in CACNA1A and PPP2R1A did not have a history of developmental conditions or neurologic disorders, including seizures.
3.3. Exploratory pathway and gene ontology analyses
Using the same list of 13 genes harboring de novo damaging variants in the anxiety disorder parent-child trios, we conducted exploratory analyses to identify shared underlying pathways (Table S3). Several biological process ontology-based sets were identified as enriched among this list of genes with the top being “neuron part” (q-value=.01) and “glutamatergic synapse” (q-value=0.01). Our list of genes with damaging variants was also enriched for several canonical pathways, including long-term depression, 16p11.2 proximal deletion syndrome, integration of energy metabolism, Wnt signaling pathway, dopaminergic synapse, mesodermal commitment, and Hippo signaling pathway.
4. DISCUSSION
By whole-exome sequencing DNA from parent-child trios, this study demonstrates for the first time a significant enrichment of rare de novo damaging variants in children with anxiety disorders compared to controls. The identification of these damaging de novo variants demonstrates a previously unexplored entry point into understanding the neurobiology of this common class of childhood psychiatric conditions, in a manner similar to what has been successful in autism spectrum disorder, Tourette’s disorder, and obsessive-compulsive disorder.
Our results suggest an approximate two-fold enrichment of rare de novo damaging variants in anxiety disorder cases compared to controls (rate ratio 1.97, 95% CI 1.11–3.34, p=0.03). Narrowing to ultra-rare de novo damaging variants that are absent in gnomAD further strengthens this statistical finding and increases the observed effect size (rate ratio 2.59, 95% CI 1.35–4.70, p=0.008). Given the wide confidence intervals, caution is warranted in interpreting these findings. Future work in larger anxiety disorder cohorts is needed to refine these estimates. Nonetheless, it is important to note that our observed enrichment of de novo LGD and Mis-D variants in childhood anxiety disorders is of a similar magnitude to previous whole-exome DNA sequencing studies in other childhood neuropsychiatric conditions, including autism (Satterstrom et al., 2020), Tourette’s disorder (S. Wang et al., 2018), and obsessive-compulsive disorder (Cappi et al., 2020; Halvorsen et al., 2021). This provides evidence that DNA sequencing studies focused on childhood anxiety disorders have the potential to identify novel high-confidence risk genes.
We identified an enrichment of de novo damaging variants in children with a spectrum of anxiety disorders, including generalized anxiety disorder, separation anxiety disorder, social phobia, and specific phobia. It is well known that these conditions are the most common in children, and previous work suggests shared genetic etiology (Tambs et al., 2009). Therefore, de novo variants identified across the spectrum of these anxiety disorders may provide common insight into neurogenetic underpinnings of these conditions.
We identified de novo damaging variants in 2 genes, CACNA1A and PPP2R1 that met criteria for being high-confidence risk genes (FDR<0.05) for other neurodevelopmental disorders (Zhao et al., 2020). CACNA1A encodes a voltage gated calcium channel subunit and has been identified as a risk gene in de novo sequencing studies of developmental disorders (Kaplanis et al., 2020) and epileptic encephalopathies (Epi, 2016). CACNA1A mutations also underlie familial hemiplegic migraine, episodic ataxia, and spinocerebellar ataxia, and individuals with these mutations often have psychiatric co-morbidities, including anxiety (Indelicato et al., 2019). PPP2R1A is a regulatory subunit of protein phosphatase-2A and has been identified as a risk gene in de novo sequencing studies of developmental disorders (Deciphering Developmental Disorders, 2015; Kaplanis et al., 2020) and has been specifically associated with intellectual disability with a range of severity (Houge et al., 2015; Lenaerts et al., 2021). Here, we identified a de novo CACNA1A frameshift indel in a 12-year-old girl with specific phobia and a de novo PPP2R1A predicted damaging missense mutation (MPC score 2.92) in a 14-year-old girl with social phobia, generalized anxiety disorder, and dysthymia. Neither of these girls are known to have other neurologic or neurodevelopmental conditions. These examples underscore the potential for pleotropic effects in which genes with de novo damaging variants lead to different clinical manifestations in unrelated individuals. Further mechanistic studies are needed to understand how these genes may lead to anxiety and other neuropsychiatric conditions.
We also conducted exploratory pathway analyses using genes with rare de novo damaging mutations in the anxiety disorder trios. In interpreting these results, it is important to keep in mind that many of these genes may not be true risk genes for childhood anxiety disorders. Nonetheless, we observed significant enrichment for several pathways and biological processes with the top gene ontology-based sets being “neuron part” and “glutamatergic synapse” (Table S3). Indeed, recent findings highlight the important role of glutamate neurotransmission in anxiety, with potential promising implications for novel treatments that modulate the glutamate system (Murrough, Yaqubi, Sayed, & Charney, 2015; Nasir et al., 2020).
Some limitations should be considered regarding this study. In comparing mutation rates, the ideal controls would be sequenced simultaneously with cases and would be assessed for anxiety disorders. In the interest of sequencing more anxiety disorder parent-child trios, this study used controls previously sequenced using similar methods that scored in the normal range on the anxiety problems subscale of the CBCL or ABCL. The fact that anxiety disorders are fairly common and can onset at any time during development could limit our observed enrichment of de novo variants in the case-control burden analysis. We attempted to minimize batch effects by focusing on the intersection of the capture platforms as successfully done in other whole-exome DNA sequencing studies using these controls (Cappi et al., 2020; S. Wang et al., 2018; Willsey et al., 2017). Another limitation is the focus on exonic regions as it is possible that important genomic changes occur in non-coding regions. However, coding variants may have a larger effect, and the biological relevance of most non-coding variation is unclear. Whole-exome DNA sequencing is also currently four-fold cheaper than whole-genome sequencing. This supports our decision to focus on whole-exome DNA sequencing in this first parent-child trio study in childhood anxiety disorders. Future work may take advantage of whole-genome DNA sequencing to assess the role of non-coding rare variation in conjunction with coding variation. Finally, our exploratory pathway analysis is preliminary and needs replication as high-confidence risk genes are identified for anxiety disorders.
Despite these limitations, our results are important because they demonstrate for the first time an enrichment of rare de novo damaging variants in children with anxiety disorders compared with controls, which reinforces the value of DNA sequencing studies in larger anxiety cohorts to identify high-confidence risk genes for childhood anxiety disorders. Discovering risk genes with large effect sizes would change the field of anxiety genetics by allowing for studies in model systems to disentangle the underlying biology and potentially identify novel treatment targets to improve interventions for patients and families.
5. CONCLUSION
In summary, rare de novo likely gene disrupting and predicted damaging missense genetic variants are enriched in probands with anxiety disorders compared to controls. These de novo damaging variants occur in individuals with a variety of childhood anxiety disorders, impact genes that have been associated with other neuropsychiatric conditions, and are enriched for canonical pathways. Our findings demonstrate a new path towards identifying risk genes and biological mechanisms in childhood anxiety disorders through de novo variant detection, which had previously been unexplored. Our results underscore the discovery potential of applying this approach in larger anxiety disorder cohorts.
Supplementary Material
Figure S1: Principal component analysis
Table S3: Pathway analysis
Table S2: De novo variants
Table S1: Quality control metrics
Acknowledgement:
We would like to thank all of the families who contributed to this research study. This study was supported by funds from the Yale Child Study Center Faculty Development Award (T.V.F.), the Yale Center for Mendelian Genomics (UM1HG006504) from the National Human Genome Research Institute as well as National Institute of Mental Health grants R01MH119299 (W.K.S.), R21MH113946 (E.R.L.), R61MH115113-01A1 (E.R.L.), and R01MH114927 (T.V.F.). E.O. was supported by National Institute of Mental Health grants R25MH077823 and T32MH018268. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Footnotes
Conflicts of Interest:
The authors declare that there are no conflicts of interest.
Data Availability Statement:
The data that supports the findings in this study are available in the supplementary material and also available upon request from the authors.
References:
- Beesdo K, Knappe S, & Pine DS (2009). Anxiety and anxiety disorders in children and adolescents: developmental issues and implications for DSM-V. Psychiatr Clin North Am, 32(3), 483–524. doi: 10.1016/j.psc.2009.06.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cappi C, Oliphant ME, Peter Z, Zai G, Conceicao do Rosario M, Sullivan CAW, … Fernandez TV (2020). De novo damaging DNA coding mutations are associated with obsessive-compulsive disorder and overlap with Tourette’s disorder and autism. Biol Psychiatry, 87(12), 1035–1044. doi: 10.1016/j.biopsych.2019.09.029 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Deciphering Developmental Disorders, S. (2015). Large-scale discovery of novel genetic causes of developmental disorders. Nature, 519(7542), 223–228. doi: 10.1038/nature14135 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Epi KC (2016). De novo mutations in SLC1A2 and CACNA1A are important causes of epileptic encephalopathies. Am J Hum Genet, 99(2), 287–298. doi: 10.1016/j.ajhg.2016.06.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Feliciano P, Zhou X, Astrovskaya I, Turner TN, Wang T, Brueggeman L, … Chung WK (2019). Exome sequencing of 457 autism families recruited online provides evidence for autism risk genes. NPJ Genom Med, 4, 19. doi: 10.1038/s41525-019-0093-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fischbach GD, & Lord C (2010). The Simons Simplex Collection: a resource for identification of autism genetic risk factors. Neuron, 68(2), 192–195. doi: 10.1016/j.neuron.2010.10.006 [DOI] [PubMed] [Google Scholar]
- Franco X, Saavedra LM, & Silverman WK (2007). External validation of comorbid patterns of anxiety disorders in children and adolescents. J Anxiety Disord, 21(5), 717–729. doi: 10.1016/j.janxdis.2006.10.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Halvorsen M, Samuels J, Wang Y, Greenberg BD, Fyer AJ, McCracken JT, … Goldstein DB (2021). Exome sequencing in obsessive-compulsive disorder reveals a burden of rare damaging coding variants. Nat Neurosci, 24(8), 1071–1076. doi: 10.1038/s41593-021-00876-8 [DOI] [PubMed] [Google Scholar]
- Herwig R, Hardt C, Lienhard M, & Kamburov A (2016). Analyzing and interpreting genome data at the network level with ConsensusPathDB. Nat Protoc, 11(10), 1889–1907. doi: 10.1038/nprot.2016.117 [DOI] [PubMed] [Google Scholar]
- Houge G, Haesen D, Vissers LE, Mehta S, Parker MJ, Wright M, … Janssens V (2015). B56delta-related protein phosphatase 2A dysfunction identified in patients with intellectual disability. J Clin Invest, 125(8), 3051–3062. doi: 10.1172/JCI79860 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Indelicato E, Nachbauer W, Karner E, Eigentler A, Wagner M, Unterberger I, … Boesch S (2019). The neuropsychiatric phenotype in CACNA1A mutations: a retrospective single center study and review of the literature. Eur J Neurol, 26(1), 66–e67. doi: 10.1111/ene.13765 [DOI] [PubMed] [Google Scholar]
- Hettema John M., M. C. N., Kendler Kenneth S.. (2001). A review and meta-analysis of the genetic epidemiology of anxiety disorders. Am J Psychiatry, 158(10), 1568–1578. [DOI] [PubMed] [Google Scholar]
- Kaplanis J, Samocha KE, Wiel L, Zhang Z, Arvai KJ, Eberhardt RY, … Retterer K (2020). Evidence for 28 genetic disorders discovered by combining healthcare and research data. Nature, 586(7831), 757–762. doi: 10.1038/s41586-020-2832-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Karczewski KJ, Francioli LC, Tiao G, Cummings BB, Alfoldi J, Wang Q, … MacArthur DG (2020). The mutational constraint spectrum quantified from variation in 141,456 humans. Nature, 581(7809), 434–443. doi: 10.1038/s41586-020-2308-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kendler KS, Gardner CO, Annas P, & Lichtenstein P (2008). The development of fears from early adolesence to young adulthood: a multivariate study. Psychol Med, 38(12), 1759–1769. doi: 10.1017/S0033291708002936 [DOI] [PubMed] [Google Scholar]
- Kessler RC, Petukhova M, Sampson NA, Zaslavsky AM, & Wittchen HU (2012). Twelve-month and lifetime prevalence and lifetime morbid risk of anxiety and mood disorders in the United States. Int J Methods Psychiatr Res, 21(3), 169–184. doi: 10.1002/mpr.1359 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Knepley MJ, Kendall PC, & Carper MM (2019). An Analysis of the Child Behavior Checklist Anxiety Problems Scale’s Predictive Capabilities. J Psychopathol Behav Assess, 41(2), 249–256. doi: 10.1007/s10862-019-09722-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lawrence PJ, Murayama K, & Creswell C (2019). Systematic review and meta-analysis: anxiety and depressive disorders in offspring of parents with anxiety disorders. J Am Acad Child Adolesc Psychiatry, 58(1), 46–60. doi: 10.1016/j.jaac.2018.07.898 [DOI] [PubMed] [Google Scholar]
- Lenaerts L, Reynhout S, Verbinnen I, Laumonnier F, Toutain A, Bonnet-Brilhault F, … Janssens V (2021). The broad phenotypic spectrum of PPP2R1A-related neurodevelopmental disorders correlates with the degree of biochemical dysfunction. Genet Med, 23(2), 352–362. doi: 10.1038/s41436-020-00981-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Levey DF, Gelernter J, Polimanti R, Zhou H, Cheng Z, Aslan M, … Stein MB (2020). Reproducible genetic risk loci for anxiety: results from approximately 200,000 participants in the million veteran program. Am J Psychiatry, 177(3), 223–232. doi: 10.1176/appi.ajp.2019.19030256 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Manichaikul A, Mychaleckyj JC, Rich SS, Daly K, Sale M, & Chen WM (2010). Robust relationship inference in genome-wide association studies. Bioinformatics, 26(22), 2867–2873. doi: 10.1093/bioinformatics/btq559 [DOI] [PMC free article] [PubMed] [Google Scholar]
- McKenna A, Hanna M, Banks E, Sivachenko A, Cibulskis K, Kernytsky A, … DePristo MA (2010). The Genome Analysis Toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res, 20(9), 1297–1303. doi: 10.1101/gr.107524.110 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Meier SM, Trontti K, Purves KL, Als TD, Grove J, Laine M, … Mors O (2019). Genetic variants associated with anxiety and stress-related disorders: a genome-wide association study and mouse-model study. JAMA Psychiatry. doi: 10.1001/jamapsychiatry.2019.1119 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Merikangas KR, He JP, Burstein M, Swanson SA, Avenevoli S, Cui L, … Swendsen J (2010). Lifetime prevalence of mental disorders in U.S. adolescents: results from the National Comorbidity Survey Replication--Adolescent Supplement (NCS-A). J Am Acad Child Adolesc Psychiatry, 49(10), 980–989. doi: 10.1016/j.jaac.2010.05.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Murrough JW, Yaqubi S, Sayed S, & Charney DS (2015). Emerging drugs for the treatment of anxiety. Expert Opin Emerg Drugs, 20(3), 393–406. doi: 10.1517/14728214.2015.1049996 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Nasir M, Trujillo D, Levine J, Dwyer JB, Rupp ZW, & Bloch MH (2020). Glutamate systems in DSM-5 anxiety disorders: their role and a review of glutamate and GABA psychopharmacology. Front Psychiatry, 11, 548505. doi: 10.3389/fpsyt.2020.548505 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Otowa T, Hek K, Lee M, Byrne EM, Mirza SS, Nivard MG, … Hettema JM (2016). Meta-analysis of genome-wide association studies of anxiety disorders. Mol Psychiatry, 21(10), 1391–1399. doi: 10.1038/mp.2015.197 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Purves KL, Coleman JRI, Meier SM, Rayner C, Davis KAS, Cheesman R, … Eley TC (2019). A major role for common genetic variation in anxiety disorders. Mol Psychiatry. doi: 10.1038/s41380-019-0559-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Samocha KE, Kosmicki JA, Karczewski KJ, O’Donnell-Luria AH, Pierce-Hoffman E, MacArthur DG, … Daly MJ (2017). Regional missense constraint improves variant deleteriousness prediction. bioRxiv. doi: 10.1101/148353 [DOI] [Google Scholar]
- Satterstrom FK, Kosmicki JA, Wang J, Breen MS, De Rubeis S, An JY, … Buxbaum JD (2020). Large-scale exome sequencing study implicates both developmental and functional changes in the neurobiology of autism. Cell, 180(3), 568–584 e523. doi: 10.1016/j.cell.2019.12.036 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Scaini S, Belotti R, & Ogliari A (2014). Genetic and environmental contributions to social anxiety across different ages: a meta-analytic approach to twin data. J Anxiety Disord, 28(7), 650–656. doi: 10.1016/j.janxdis.2014.07.002 [DOI] [PubMed] [Google Scholar]
- Scaini S, Ogliari A, Eley TC, Zavos HM, & Battaglia M (2012). Genetic and environmental contributions to separation anxiety: a meta-analytic approach to twin data. Depress Anxiety, 29(9), 754–761. doi: 10.1002/da.21941 [DOI] [PubMed] [Google Scholar]
- Silverman WK, Saaverdram LM, & Pina AA (2001). Test-retest reliability of anxiety symptoms and diagnoses with the Anxiety Disorders Interview Schedule for DSM-IV: child and parent versions. Journal of the American Academy of Child & Adolescent Psychiatry, 40(8), 937–944. [DOI] [PubMed] [Google Scholar]
- Smoller JW (2016). The genetics of stress-related disorders: PTSD, depression, and anxiety disorders. Neuropsychopharmacology, 41(1), 297–319. doi: 10.1038/npp.2015.266 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tambs K, Czajkowsky N, Roysamb E, Neale MC, Reichborn-Kjennerud T, Aggen SH, … Kendler KS (2009). Structure of genetic and environmental risk factors for dimensional representations of DSM-IV anxiety disorders. Br J Psychiatry, 195(4), 301–307. doi: 10.1192/bjp.bp.108.059485 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang K, Li M, & Hakonarson H (2010). ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res, 38(16), e164. doi: 10.1093/nar/gkq603 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang S, Mandell JD, Kumar Y, Sun N, Morris MT, Arbelaez J, … State MW (2018). De novo sequence and copy number variants are strongly associated with tourette disorder and implicate cell polarity in pathogenesis. Cell Rep, 24(13), 3441–3454 e3412. doi: 10.1016/j.celrep.2018.08.082 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Waszczuk MA, Zavos HM, Gregory AM, & Eley TC (2014). The phenotypic and genetic structure of depression and anxiety disorder symptoms in childhood, adolescence, and young adulthood. JAMA Psychiatry, 71(8), 905–916. doi: 10.1001/jamapsychiatry.2014.655 [DOI] [PubMed] [Google Scholar]
- Willsey AJ, Fernandez TV, Yu D, King RA, Dietrich A, Xing J, … Heiman GA (2017). De novo coding variants are strongly associated with Tourette disorder. Neuron, 94(3), 486–499 e489. doi: 10.1016/j.neuron.2017.04.024 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhao G, Li K, Li B, Wang Z, Fang Z, Wang X, … Li J (2020). Gene4Denovo: an integrated database and analytic platform for de novo mutations in humans. Nucleic Acids Res, 48(D1), D913–D926. doi: 10.1093/nar/gkz923 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Figure S1: Principal component analysis
Table S3: Pathway analysis
Table S2: De novo variants
Table S1: Quality control metrics
Data Availability Statement
The data that supports the findings in this study are available in the supplementary material and also available upon request from the authors.