RNA sequencing (RNA-seq) is a powerful tool in diagnosing Mendelian disorders, but the optimized sequencing depth for this technology has not yet been determined. In a new study published in the American Journal of Human Genetics, researchers at Baylor College of Medicine's Medical Genetics Multiomics Laboratory (MGML) show the utility and benefit of ultra-deep RNA sequencing in clinical diagnostics.
Most current RNA-seq diagnostic protocols use moderate sequencing depths (about 50-150 million reads) that may not detect low-abundance transcripts and rare splicing events that are critical to clinical interpretation of gene variants of uncertain significance. The Baylor team wanted to investigate the diagnostic utility of ultra-high depth (up to 1 billion reads) RNA-seq.
"Using the Ultima Genomics ultra-high depth RNA-seq platform, we substantially increased the detection of lowly expressed genes and transcripts," said corresponding author Dr. Pengfei Liu, MGML director and associate professor of molecular and human genetics at Baylor College of Medicine. "With ultra-deep sequencing, we can evaluate many more low-expression genes that would be missed when sequenced at traditional sequencing depths."
The ability to detect low-expression genetic variants is especially important in the clinical diagnostic setting. Clinical diagnostic labs most frequently test blood and skin samples. However, genes causing developmental and neurological disorders may not be strongly expressed in blood and skin cells.
"Our study shows that if you can sequence blood samples to extremely high depths, you can capture those genes traditionally thought to be tissue specific," said Liu.
The data in this study can be used to guide future RNA-seq research. The team used the deep RNA-seq data to create an online resource that estimates the required sequencing depth needed to achieve a genetic diagnosis. The tool also can be used to predict pathogenic abnormal gene splicing from deeply sequenced healthy people data.
Next, the researchers will perform clinical validation for ultra-deep RNA-seq and plan for a clinical test based on the findings of this study. "In the MGML, we are leaders in translating new genomic technologies into real-world clinical practice," Liu said. "We continue to evaluate new technology that can help improve diagnostic rates for our patients."
Other authors who contributed to this work include: Sen Zhao, Jefferson C. Sinson, Shenglan Li, Jill A. Rosenfeld, Gladys Zapata, Kristina Macakova, Mezthly Pena, Becky Maywald, Kim C. Worley, Lindsay Burrage, Monika Weisz Hubshman, Shamika Ketkar, William Craigen, Lisa Emrick, Tyson Clark, Gila Yanai Lithwick, Zohar Shipony, Christine Eng and Brendan Lee. They are affiliated with Baylor College of Medicine, Ultima Genomics and Baylor Genetics. This work is conducted in collaboration with the Undiagnosed Diseases Network.
This work was supported by the National Institute of Neurological Disorders and Stroke (U01HG007709 and U01HG007942) and the National Human Genome Research Institute (R35HG011311).
