Researchers from Baylor College of Medicine have demonstrated that using ultra-deep RNA sequencing can dramatically improve the diagnosis of genetic diseases. Traditional RNA sequencing typically captures about 50 to 150 million reads per sample, which is often enough to study common gene expression patterns. However, this level of coverage can miss important but rare signals, such as low-abundance transcripts or subtle splicing errors, that are critical for understanding how certain mutations cause disease.

By pushing sequencing depth up to 1 billion reads per sample, the researchers showed that many previously undetectable gene variants could now be identified. These include splicing abnormalities linked to disease-causing variants that standard methods failed to capture. In two patient cases, the team found that pathogenic RNA changes were invisible at 50 million reads but became apparent when sequencing depth reached 200 million, and even clearer at 1 billion reads.

The researchers used the Ultima sequencing platform to validate their deep RNA sequencing approach across four types of human tissues. They then developed a valuable tool called MRSD-deep, which helps scientists determine how much sequencing depth is needed to confidently detect specific genes or splicing events. This tool provides gene- and junction-level guidelines that laboratories can use to plan future experiments more efficiently.

In addition, the team created an expanded splicing reference database from deep RNA sequencing data on fibroblasts. This resource captured many low-abundance splicing variations that were previously missed, helping improve the interpretation of variants of uncertain significance (VUSs).

Overall, this work highlights how ultra-deep RNA sequencing can reveal critical biological insights that standard methods overlook, bringing scientists closer to accurate diagnoses and better understanding of rare Mendelian disorders.