UPDhmm: detecting uniparental disomy from NGS trio data

Area of Science:

• Genetics
• Genomic analysis
• Chromosomal abnormalities

Background:

• Uniparental disomies (UPDs) are chromosomal alterations where both chromosome copies originate from one parent.
• UPDs can result from meiotic or mitotic segregation errors and are linked to congenital and acquired diseases.
• Current sequence-based UPD detection methods lack sensitivity for small events and struggle with consanguinity.

Purpose of the Study:

• To introduce UPDhmm, a novel Hidden Markov Model (HMM)-based tool for detecting UPDs.
• To enhance the accuracy and sensitivity of UPD detection using trio-based sequence data.
• To demonstrate the clinical utility of UPDhmm in identifying UPDs associated with autism spectrum disorder.

Main Methods:

• UPDhmm utilizes trio-based (proband and parents) exome or genome sequence data.
• It employs a Hidden Markov Model (HMM) to predict inheritance patterns, distinguishing between Mendelian inheritance and UPD events.
• The tool was validated using simulated data from the 1000-Genomes project and applied to the Simons Simplex Collection.

Main Results:

• UPDhmm demonstrated superior performance over existing methods in detecting simulated UPD events in both exome and genome data.
• Application to ~2400 families revealed UPD events in two individuals with autism spectrum disorder.
• Identified UPDs include a paternal isodisomy of chromosome 8 and a maternal heterodisomy of chromosome 22.

Conclusions:

• UPDhmm offers a sensitive and precise method for UPD detection using trio sequencing.
• The tool can be integrated into clinical genomic analysis pipelines for improved disease diagnosis.
• Detected UPD events highlight their potential role as genetic causes of autism spectrum disorder.