| Background:Autism spectrum disorder（ASD）is a group of neurodevelopmental disorders with high clinical and etiological heterogeneity that seriously affect children’s health.With the exception of a small number of ASDs caused solely by genetic factors,the etiology of most ASDs is complex and may involve the interaction of genetic variants（including common and rare variants）and environmental factors.It has been found that high-confidence ASD pathogenic or risk genes are mainly involved in biological processes such as gene expression regulation.Stress granule（SG）is a membraneless organelle formed by cells,including neurons,in the face of external environmental stimuli.Its dynamic changes are one of the main defense methods of cells in response to environmental stress.The formation of SG can cause the arrest of intracellular protein translation,thereby affecting the regulation of gene expression.Studies have shown that dysfunction of SG-related pathways is associated with a variety of neurodegenerative diseases.In addition,mutations in key regulatory genes of multiple SGs,such as CSDE1,DDX3X,CAPRIN1,and FMR1,can lead to ASD and related neurodevelopmental disorders.However,the genetic regulatory relationship and mechanism of SG-related genes and ASD risk are still unclear.Objective:This study used the methods of genomics and genetic statistics to explore the relationship between common and rare variants of SG-related genes and the risk of ASD,and provided a basis for further research on the mechanism of SG genes involved in the pathogenesis of ASD and the possible interaction between genes and the environment.Methods:The data in this study were obtained from the whole genome sequencing data of 2023 ASD families in the SPARK data set and 2344ASD families in the SSC data set provided by the SFARI organization in the United States.GATK was used to train the variant site quality scoring model,filtering out variants of low-quality genotypes（GQ<20 and DP<15）,and excluding low-complexity regions and those that did not meet Hardy-Weinberg equilibrium(P<10-6).Mutations.PLINK was used to exclude variant sites with a missing rate greater than 0.8 in the total sample and samples with a missing rate greater than 0.8 in the total variation.Principal component analysis and population structure analysis were performed using PLINK.Kinship was estimated using KING and families with unmatched kinship were excluded.Single locus-based association analysis was performed for common variants（MAF>0.01）using PLINK’s Transmission Disequilibrium Test（TDT）.Gene-by-gene TDT analysis of rare（MAF<0.001）coding region variants was performed using RV-TDT.The R package cluster Profiler was used to perform enrichment analysis and pathway analysis on candidate genes obtained by association analysis.Results:（1）Data quality control analysis found that the number of C>T and G>A variant types in SNPs was the largest,followed by A>G and T>C variant types,and the Ts/Tv ratio was 2.1;most INDEL sites were concentrated in 10 bp,and the number of insertion variants is slightly less than the number of deletion variants;the average sequencing depth of variants after data filtering is greater than 30×,and all variants in the population conform to the Hardy-Weinberg law.（2）The population structure in the SPARK and SSC datasets is relatively scattered,and can be roughly divided into Asian,European,and African.The kinship relationships of all families in the SPARK dataset are in line with the family relationships originally provided by the sample database,and there are 7family relationships in the SSC dataset that do not conform to the family relationships originally provided by the sample database.（3）The number of common variances included in the TDT analysis of the SPARK dataset was 154,937,and the inflation factor（λ）was 1.002.Twenty variants in 12SG genes including RELN,SPART,and RBFOX1 reached the significance threshold after Bonferroni correction(P<3.23×10-7).The number of common variants included in the TDT analysis for the SSC dataset was144,668,with aλof 1.038.Nineteen variants in 12 SG genes,including RELN,SPART,and KANK2,reached the significance threshold after Bonferroni correction(P<3.46×10-7).（4）In the TDT analysis of rare loss-of-function variants,CNOT8 exceeded the significant threshold in both SPARK and SSC datasets.In the TDT analysis of rare missense variants,TNKS1BP1 exceeded the threshold for significant difference in both the SPARK and SSC datasets.（5）Candidate genes found by association analysis are mainly involved in biological processes such as m RNA metabolism,transcription factor complex binding and ribonucleoprotein particle formation.Conclusion:In this study,84 candidate genes for ASD were found in843 SG genes,including RELN,RBFOX1,SPART,CNOT8,TNKS1BP1,etc.,through association analysis of common variation loci and rare variation based on genes.These results suggest that SG gene variation is associated with the risk of ASD,suggesting that changes in SG function may be associated with the pathogenesis of ASD.33 figures,14 tables and 74 references... |
PDF Full Text Request