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

Publisher: MDPI AG Country of Publication: Switzerland NLM ID: 101606588 Publication Model: Electronic Cited Medium: Print ISSN: 2077-0383 (Print) Linking ISSN: 20770383 NLM ISO Abbreviation: J Clin Med Subsets: PubMed not MEDLINE

Original Publication: Basel, Switzerland : MDPI AG, [2012]-

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.

Curr Oncol Rep. 2022 Jun;24(6):671-686. (PMID: 35230594)
JAMA Oncol. 2023 Sep 1;9(9):1283-1290. (PMID: 37440251)
Pediatr Blood Cancer. 2010 Aug;55(2):304-8. (PMID: 20583218)
Cancers (Basel). 2022 Sep 26;14(19):. (PMID: 36230609)
Pharmaceutics. 2021 May 26;13(6):. (PMID: 34073456)
Ann Oncol. 2017 Feb 1;28(2):386-392. (PMID: 28426102)
Nat Biomed Eng. 2018 Oct;2(10):749-760. (PMID: 31001455)
Pediatr Blood Cancer. 2022 Nov;69(11):e29937. (PMID: 36083863)
Biology (Basel). 2023 Sep 21;12(9):. (PMID: 37759668)
Int J Lab Hematol. 2019 Dec;41(6):717-725. (PMID: 31498973)
Blood. 2014 Aug 14;124(7):1056-61. (PMID: 24904116)
Cancer Inform. 2019 Mar 15;18:1176935119835544. (PMID: 30911218)
Inf Process Med Imaging. 2023 Jun;13939:208-221. (PMID: 38680427)
Nat Mach Intell. 2020 Jan;2(1):56-67. (PMID: 32607472)
J Clin Med. 2015 Jan 09;4(1):127-49. (PMID: 26237023)
Pediatr Blood Cancer. 2012 May;58(5):701-7. (PMID: 21793184)
Med Biol Eng Comput. 2020 Nov;58(11):2631-2640. (PMID: 32840766)
Pediatr Blood Cancer. 2019 Jun;66(6):e27701. (PMID: 30848067)
Lancet Haematol. 2022 Sep;9(9):e678-e688. (PMID: 35870472)
J Rheumatol. 2017 Oct;44(10):1484-1486. (PMID: 28765255)
PLOS Digit Health. 2022 Feb 17;1(2):e0000014. (PMID: 36812540)
Big Data Cogn Comput. 2022 Mar;6(1):. (PMID: 35936510)
Nat Commun. 2018 Jan 3;9(1):42. (PMID: 29298978)
Genet Epidemiol. 2019 Oct;43(7):844-863. (PMID: 31407831)
Proc IEEE Int Conf Acoust Speech Signal Process. 2023 Jun;2023:. (PMID: 38682049)
JAMA Oncol. 2018 Mar 1;4(3):403-404. (PMID: 29392271)
Pathol Oncol Res. 2022 Apr 05;28:1610096. (PMID: 35449729)
Front Immunol. 2021 Feb 19;11:627662. (PMID: 33679722)
iScience. 2020 Jan 24;23(1):100780. (PMID: 31918046)
Asian Pac J Cancer Prev. 2020 Nov 01;21(11):3211-3219. (PMID: 33247677)
Bioengineering (Basel). 2023 Aug 02;10(8):. (PMID: 37627803)
Cell. 2020 Apr 2;181(1):92-101. (PMID: 32243801)
Brain Sci. 2021 Jul 23;11(8):. (PMID: 34439596)
Lancet Digit Health. 2021 Nov;3(11):e745-e750. (PMID: 34711379)
Cureus. 2023 Oct 11;15(10):e46837. (PMID: 37954717)
Pediatr Blood Cancer. 2018 May;65(5):e26952. (PMID: 29319209)
Scand J Infect Dis. 2003;35(11-12):851-7. (PMID: 14723361)
Int J Mol Sci. 2023 Aug 02;24(15):. (PMID: 37569714)
Cancer Treat Res. 2011;157:339-62. (PMID: 21052965)
Leuk Lymphoma. 2021 Apr;62(4):899-908. (PMID: 33258395)
Pediatr Blood Cancer. 2020 Dec;67(12):e28673. (PMID: 32918533)
Comput Struct Biotechnol J. 2023 Mar 29;21:2454-2470. (PMID: 37077177)
J Pediatr Hematol Oncol. 2022 Aug 1;44(6):273-279. (PMID: 35700365)
Pediatr Hematol Oncol. 2001 Apr-May;18(3):167-72. (PMID: 11293283)
Oncotarget. 2016 Dec 13;7(50):83767-83774. (PMID: 27590519)
Leuk Lymphoma. 2019 Sep;60(9):2104-2114. (PMID: 30774019)
JCO Clin Cancer Inform. 2021 Dec;5:1208-1219. (PMID: 34910588)
Arthritis Res Ther. 2022 Jan 3;24(1):5. (PMID: 34980230)
Jpn J Clin Oncol. 1984 Mar;14(1):31-40. (PMID: 6231391)
Eur J Oncol Nurs. 2015 Feb;19(1):13-7. (PMID: 25227459)
Eur J Clin Nutr. 2013 Oct;67(10):1056-9. (PMID: 23963274)
J Pediatr Hematol Oncol. 2023 Mar 1;45(2):e154-e160. (PMID: 36715999)
Front Bioeng Biotechnol. 2019 Jul 03;7:156. (PMID: 31334227)
Proc Inst Mech Eng H. 2020 Oct;234(10):1051-1069. (PMID: 32633668)
Orphanet J Rare Dis. 2018 Jan 19;13(1):14. (PMID: 29351763)
J Pediatr Hematol Oncol. 2016 Aug;38(6):423-8. (PMID: 26925709)
Rev Bras Hematol Hemoter. 2013;35(1):39-43. (PMID: 23580883)
Int ACM SIGIR Conf Res Dev Inf Retr. 2023 Jul;2023:2913-2923. (PMID: 38690157)

R21 CA232249 United States CA NCI NIH HHS; R35 GM152245 United States GM NIGMS NIH HHS; R21CA232249 United States NH NIH HHS

Keywords: artificial intelligence; infection; machine learning; natural language processing; pediatric leukemia

Date Created: 20240328 Latest Revision: 20241110

20241110

PMC10970787

10.3390/jcm13061788

38542012