SUsPECT: a pipeline for variant effect prediction based on custom long-read transcriptomes for improved clinical...

Area of Science:

• Genomics
• Transcriptomics
• Bioinformatics

Background:

• Incomplete human transcriptome knowledge hinders disease-causing variant detection, especially for condition-specific transcripts missing from reference sets.
• Novel transcripts, crucial for genetic diagnoses, are often absent in standard reference transcriptomes like Ensembl/GENCODE and RefSeq.

Purpose of the Study:

• To introduce SUsPECT (Solving Unsolved Patient Exomes/gEnomes using Custom Transcriptomes), a pipeline for predicting variant impact on custom transcript sets.
• To enhance the detection and prioritization of disease-causing genetic variants by incorporating novel or condition-specific transcriptomic data.

Main Methods:

• SUsPECT utilizes the Ensembl Variant Effect Predictor (VEP) to analyze custom transcriptomes, including those from long-read RNA sequencing.
• The pipeline predicts functional consequences and deleteriousness scores for missense variants within novel open reading frames.

Main Results:

• SUsPECT identified potential pathogenic mechanisms for variants in ClinVar missed by standard reference annotations.
• Annotating with a custom transcriptome from stimulated immune cells revealed an enrichment of severe molecular consequences for immune-related variants compared to reference annotations.

Conclusions:

• SUsPECT provides critical information for prioritizing potentially disease-causing variants, improving genetic diagnostics.
• The pipeline's utility is expected to grow with the increasing availability of long-read RNA sequencing data.