A new collaborative research article between Dr. Byung Yoon Choi’s research team from Seoul National University Bundang Hospital and 3billion was recently published in Scientific Report.1

In the study, researchers concluded that the use of an automated prioritization system reduced the time for diagnosis from 2 hours to 5 minutes, and diagnostic results were 97.7% concordant with those of clinical genetic experts. Let’s look into the details of the research and the future of automated variant prioritization.

Clinical study of sensorineural hearing loss diseases using an automated variant prioritization system

The research team of Dr. Byung Yoon Choi, an expert in genetics and hearing loss, conducted the study on 263 patients with mild to severe sensorineural hearing loss(SNHL).

SNHL is a disease where more than 50% of cases are attributed to genetic factors and 95% of them follow Mendelian inheritance patterns; thus, genetic testing is suitable for its diagnosis.

The purpose of the study is as follows:

  1. Check the accuracy of variant prioritization by the automated variant prioritization system and bioinformaticians,
  2. Check their concordance,
  3. Analyze a challenging VUS case using the automated prioritization system.

3billion's EVIDENCE was used as the automated variant prioritization system.

Diagnostic yield of over 50%, concordance of 97.7% between EVIDENCE and genetic experts

According to the study, candidate variants were found in 50.95% of the patients when the automated variant prioritization system was used, and 50.19% through manual variant prioritization of bioinformaticians.

As a result, performance of the automated variant prioritization system was found to be equivalent to or slightly better than variant prioritization by bioinformaticians, and the concordance rate of the two was very high, scoring 97.7%.

Variant prioritization of a sample using the automated variant prioritization system took 5 minutes, reducing the time taken for diagnosis from 2 hours by 24 times.

Potential for further development of automated variant prioritization systems

While it is not yet possible to conclude that the same performance will be achieved for all genetic diseases as the diagnosis rate of hearing loss is relatively high compared to other disease groups, the study successfully demonstrated the power of an automated system in variant prioritization for hearing loss diagnosis using phenotype-driven exome sequencing software in the largest cohort size to date.

In addition, this study showed the limitations of each method (automated and manual variant prioritization) and potential for further development. It is expected that the method where both a clinical geneticist takes part and an automated variant prioritization system is used will increase the accuracy and efficiency of diagnosis based on exome sequencing data.

Reference

  1. Kim, S.Y., Lee, S., Seo, G.H. et al. Powerful use of automated prioritization of candidate variants in genetic hearing loss with extreme etiologic heterogeneity. Sci Rep 11, 19476 (2021).