# Large-scale sequencing study of de novo regulatory Tandem Repeats (TRs) identifies new ASD (Autism Spectrum Disorders) candidate genes integrating gene expression mapping, brain scRNA-seq and organoid models

## TL;DR

This study identifies new autism risk genes by analyzing genetic variations in non-coding regions using advanced sequencing and brain cell data.

## Contribution

The study introduces a novel integrative approach combining de novo tandem repeats with gene expression and brain cell data to uncover ASD candidate genes.

## Key findings

- The gene ECHS1 was identified as a strong autism candidate through multiple lines of evidence.
- De novo tandem repeats in non-coding regions were linked to altered gene expression in brain cells.
- Integrating genetic and transcriptomic data improves the detection of autism risk genes missed by traditional methods.

## Abstract

In this study, we performed an integrative analysis of de novo tandem repeats (TRs) to unravel the missing heritability that may be hidden in 85,394 active cis-regulatory elements (cCREs) from ENCODE through target sequencing in a Spanish cohort of 200 ASD trios, using a robust bioinformatic pipeline. For the integrative analysis, we use data from 1,637 ASD simplex quad families from the Simons Simplex Collection (SSC). We then incorporated multiple layers of functional annotation, including predicted transcription factor (TF) binding sites, gene mapping based on physical proximity and expression correlation, pathogenicity scoring, single-cell RNA-seq data from human brain in ASD cases and controls and cortical organoid expression data.

Together, our analyses identified multiple ASD-relevant candidate genes supported by convergent lines of evidence. Notably, ECHS1 emerged as a strong candidate, affected by several de novo TRs in both the Spanish cohort and the SSC. It was also identified as the most significantly associated gene through expression-based gene mapping (T-Gene) and showed consistent differential expression in excitatory neurons of the cerebral cortex at the single-cell level along with increased expression in late-stage cortical organoids.

These findings remark the value of integrating genetic and transcriptomic information to improve the identification of potential risk genes for ASD, particularly within non-coding regions. Our approach also highlights the importance of identifying complex genetic variation, such as de novo TRs, that are typically missed in conventional exome or whole-genome analyses, and require specialized bioinformatic strategies for accurate detection and interpretation.