Using the linear references from the pangenome to discover missing autism variants

Area of Science:

• Genomics
• Neurodevelopmental Disorders
• Human Genetics

Background:

• Understanding large-effect pathogenic variation is crucial for autism spectrum disorder (ASD) etiology.
• Previous studies using short-read sequencing have limitations in detecting complex structural variants (SVs).
• Phased genome assemblies offer a more comprehensive view of genetic variation.

Purpose of the Study:

• To characterize de novo mutations, structural variants (SVs), and DNA methylation in individuals with unsolved autism cases using long-read sequencing (LRS).
• To develop and apply a workflow for prioritizing pathogenic variants by integrating autism risk genes and regulatory elements.
• To evaluate the utility of phased genomes and pangenome controls for identifying complex mutations.

Main Methods:

• Generated long-read sequencing (LRS) data for 189 individuals from 51 families with unsolved autism cases.
• Constructed phased and near-complete genome assemblies.
• Applied read- and assembly-based strategies for variant characterization, including SVs and DNA methylation, using LRS pangenome controls.

Main Results:

• Filtered >97% of common SVs using LRS pangenome controls.
• Observed no increased autosomal SV burden in probands versus siblings, but a trend toward increased X chromosome SV burden in affected females.
• Identified three pathogenic variants (TBL1XR1, MECP2, SYNGAP1) and nine candidate de novo/biallelic inherited SVs, many missed by short-read sequencing.

Conclusions:

• Phased genomes are powerful for discovering complex and pathogenic mutations associated with autism.
• LRS and pangenome approaches significantly enhance the detection of structural variants.
• The developed prioritization workflow aids in identifying clinically relevant variants for autism.