An Interpretable Machine Learning Framework for Rare Disease: A Case Study to Stratify Infection Risk in Pediatric Leukemia

Abstract

Background: Datasets on rare diseases, like pediatric acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL), have small sample sizes that hinder machine learning (ML). The objective was to develop an interpretable ML framework to elucidate actionable insights from small tabular rare disease datasets. Methods: The comprehensive framework employed optimized data imputation and sampling, supervised and unsupervised learning, and literature-based discovery (LBD). The framework was deployed to assess treatment-related infection in pediatric AML and ALL. Results: An interpretable decision tree classified the risk of infection as either “high risk” or “low risk” in pediatric ALL (n = 580) and AML (n = 132) with accuracy of ∼79%. Interpretable regression models predicted the discrete number of developed infections with a mean absolute error (MAE) of 2.26 for bacterial infections and an MAE of 1.29 for viral infections. Features that best explained the development of infection were the chemotherapy regimen, cancer cells in the central nervous system at initial diagnosis, chemotherapy course, leukemia type, Down syndrome, race, and National Cancer Institute risk classification. Finally, SemNet 2.0, an open-source LBD software that links relationships from 33+ million PubMed articles, identified additional features for the prediction of infection, like glucose, iron, neutropenia-reducing growth factors, and systemic lupus erythematosus (SLE). Conclusions: The developed ML framework enabled state-of-the-art, interpretable predictions using rare disease tabular datasets. ML model performance baselines were successfully produced to predict infection in pediatric AML and ALL.

Cross-Domain Text Mining of Pathophysiological Processes Associated with Diabetic Kidney Disease

“…Domain knowledge provides an important layer of validation. The results of SemNet 2.0 studies have been successfully evaluated in COVID-19 [34,36], Parkinson's disease [35], chronic myeloid leukemia [29], and pediatric acute leukemia [37]. In fact, the majority of repurposed drug candidates recommended by SemNet early in the pandemic advanced to successful clinical trials as adjuvant therapies [94].…”

“…The unsupervised learning ranking algorithm within SemNet 2.0 examines relationship patterns in the literature to rank cross-domain concepts with respect to the user-defined concept(s) [31]. SemNet 2.0 has been used for drug repurposing for COVID-19 [34] and Parkinson's disease [35], identifying unknown disease mechanisms of resistant hypertension following COVID-19 infection [36], predicting adverse events from chronic tyrosine kinase inhibitor therapy in chronic myeloid leukemia [29], and identifying clinical features by which to better stratify chemotherapy-related infection risk in pediatric acute leukemia [37].…”