Genetic testing concept, DNA icon, medical doctor, isolated on white. Illustrating RNA sequencing to diagnose rare diseases
Credit: NicoElNino/Getty Images

Results from the 100,000 Genomes cohort in the U.K. show that RNA sequencing has the potential to increase the number of diagnoses given to individuals with rare disease.

The researchers predict that over 20% of people in the cohort who did not achieve a diagnosis with genome or exome sequencing can achieve one by using this method.

Jenny Lord, a senior research fellow in computational biology at the University of Southampton, reported the results of this analysis at the European Society of Human Genetics conference in Glasgow this week.

“The diagnosis of rare disorders has really been revolutionized in recent years through the likes of whole exome and whole genome sequencing. But even with whole genome sequencing, about half of patients don’t get a diagnosis,” she explained.

“A lot of this has to do with the fact that our ability to generate this data has really outstripped our ability to interpret it. RNA sequencing gives us a way to get around this interpretation bottleneck, and see what’s actually being produced from the gene at the RNA level. We can look for places where a patient is different from controls, or from other patients, in terms of expression or splicing, and link that back to changes in the DNA and ultimately make diagnoses for patients.”

Lord and colleagues analyzed whole blood-based RNA sequencing data from 4400 individuals who participated in the 100,000 Genomes Project, which began in 2012. Of this group, 36% have some form of neurodevelopmental disorder, 49% are female and 70% are from White European ancestry. The people in this group previously failed to receive a diagnosis of their condition after standard whole genome sequencing.

Gene expression outliers were identified by OUTRIDER via DROP and splicing outliers were mostly identified using LeafCutterMD. Overall, an average of 5.4 expression and 5.3 splicing outliers were found across the genome of each participant tested. When these were limited to panels of relevant genes, this dropped to 0.2 expression and 0.2 splicing outliers per person tested.

For example, a splicing outlier was found in the PTEN gene, which is linked to Cowden syndrome. On the expression side, an outlier was found for the RPL5 gene, which is linked to Diamond-Blackfan anemia.

“We’ve already been able to identify many expression and splicing outliers, some of which we already think are diagnostic, and the work on this is ongoing,” explained Lord, who added that there is lots of analysis still to be done.

“Our early estimates say that we should find one of these candidate diagnostic events in about 25% of the cohort, which has the potential to make a huge impact on diagnostic rate.”

Also of Interest