Artificial intelligence (AI)-assisted exome reanalysis greatly aids in the identification of new positive cases and reduces analysis time in a clinical diagnostic laboratory

Timothy D. O'Brien, N. Eleanor Campbell, Amiee B. Potter, John H. Letaw, Arpita Kulkarni, C. Sue Richards

Research output: Contribution to journalArticlepeer-review

4 Scopus citations

Abstract

Purpose: Artificial intelligence (AI) and variant prioritization tools for genomic variant analysis are being rapidly developed for use in clinical diagnostic testing. However, their clinical utility and reliability are currently limited. Therefore, we performed a validation of a commercial AI tool (Moon) and a comprehensive reanalysis of previously collected clinical exome sequencing cases using an open-source variant prioritization tool (Exomiser) and the now-validated AI tool to test their feasibility in clinical diagnostics. Methods: A validation study of Moon was performed with 29 positive cases determined by previous manual analysis. After validation, reanalysis was performed on 80 previously manually analyzed nondiagnostic exome cases using Moon. Finally, a comparison between Moon and Exomiser was completed regarding their ability to identify previously completed positive cases and to identify new positive cases. Results: Moon correctly selected the causal variant(s) in 97% of manually analyzed positive cases and identified 7 new positive cases. Exomiser correctly identified the causal gene in 85% of positive cases and agreed with Moon by ranking the new gene in its top 10 list 43% of the time. Conclusion: The use of AI in diagnostic laboratories greatly enhances exome sequencing analysis by reducing analysis time and increasing the diagnostic rate.

Original languageEnglish (US)
Pages (from-to)192-200
Number of pages9
JournalGenetics in Medicine
Volume24
Issue number1
DOIs
StatePublished - Jan 2022

Keywords

  • Artificial intelligence
  • Exome sequencing
  • Reanalysis
  • Variant classification
  • Variant prioritization

ASJC Scopus subject areas

  • Genetics(clinical)

Fingerprint

Dive into the research topics of 'Artificial intelligence (AI)-assisted exome reanalysis greatly aids in the identification of new positive cases and reduces analysis time in a clinical diagnostic laboratory'. Together they form a unique fingerprint.

Cite this