A geno-clinical decision model for the diagnosis of myelodysplastic syndromes

The differential diagnosis of myeloid malignancies is challenging and subject to interobserver variability. We used clinical and next-generation sequencing (NGS) data to develop a machine learning model for the diagnosis of myeloid malignancies independent of bone marrow biopsy data based on a 3-institution, international cohort of patients. The model achieves high performance, with model interpretations indicating that it relies on factors similar to those used by clinicians. In addition, we describe associations between NGS findings and clinically important phenotypes and introduce the use of machine learning algorithms to elucidate clinicogenomic relationships.

Automated summary

By utilizing clinical and next-generation sequencing data, a machine learning model was successfully developed for the diagnosis of myeloid malignancies independent of bone marrow biopsy data in an international patient cohort, achieving high performance. The model interpretations suggest that it relies on factors similar to those used by clinicians, and associations between NGS findings and clinically important phenotypes were described. The use of machine learning algorithms to elucidate clinicogenomic relationships was also introduced.