Dataset Information

Development of Prediction Models Using Machine Learning Algorithms for Girls with Suspected Central Precocious Puberty: Retrospective Study.

ABSTRACT: BACKGROUND:Central precocious puberty (CPP) in girls seriously affects their physical and mental development in childhood. The method of diagnosis-gonadotropin-releasing hormone (GnRH)-stimulation test or GnRH analogue (GnRHa)-stimulation test-is expensive and makes patients uncomfortable due to the need for repeated blood sampling. OBJECTIVE:We aimed to combine multiple CPP-related features and construct machine learning models to predict response to the GnRHa-stimulation test. METHODS:In this retrospective study, we analyzed clinical and laboratory data of 1757 girls who underwent a GnRHa test in order to develop XGBoost and random forest classifiers for prediction of response to the GnRHa test. The local interpretable model-agnostic explanations (LIME) algorithm was used with the black-box classifiers to increase their interpretability. We measured sensitivity, specificity, and area under receiver operating characteristic (AUC) of the models. RESULTS:Both the XGBoost and random forest models achieved good performance in distinguishing between positive and negative responses, with the AUC ranging from 0.88 to 0.90, sensitivity ranging from 77.91% to 77.94%, and specificity ranging from 84.32% to 87.66%. Basal serum luteinizing hormone, follicle-stimulating hormone, and insulin-like growth factor-I levels were found to be the three most important factors. In the interpretable models of LIME, the abovementioned variables made high contributions to the prediction probability. CONCLUSIONS:The prediction models we developed can help diagnose CPP and may be used as a prescreening tool before the GnRHa-stimulation test.

SUBMITTER: Pan L

PROVIDER: S-EPMC6390190 | biostudies-other | 2019 Feb

REPOSITORIES: biostudies-other

ACCESS DATA

Publications

Development of Prediction Models Using Machine Learning Algorithms for Girls with Suspected Central Precocious Puberty: Retrospective Study.

Pan Liyan L Liu Guangjian G Mao Xiaojian X Li Huixian H Zhang Jiexin J Liang Huiying H Li Xiuzhen X

JMIR medical informatics 20190212 1

<h4>Background</h4>Central precocious puberty (CPP) in girls seriously affects their physical and mental development in childhood. The method of diagnosis-gonadotropin-releasing hormone (GnRH)-stimulation test or GnRH analogue (GnRHa)-stimulation test-is expensive and makes patients uncomfortable due to the need for repeated blood sampling.<h4>Objective</h4>We aimed to combine multiple CPP-related features and construct machine learning models to predict response to the GnRHa-stimulation test.<h ...[more]

PMID: 30747712

Similar Datasets

Project description:BackgroundRecent studies demonstrated that changes in DNA methylation (DNAm) and inactivation of two imprinted genes (MKRN3 and DLK1) alter the onset of female puberty. We aimed to investigate the association of DNAm profiling with the timing of human puberty analyzing the genome-wide DNAm patterns of peripheral blood leukocytes from ten female patients with central precocious puberty (CPP) and 33 healthy girls (15 pre- and 18 post-pubertal). For this purpose, we performed comparisons between the groups: pre- versus post-pubertal, CPP versus pre-pubertal, and CPP versus post-pubertal.ResultsAnalyzing the methylome changes associated with normal puberty, we identified 120 differentially methylated regions (DMRs) when comparing pre- and post-pubertal healthy girls. Most of these DMRs were hypermethylated in the pubertal group (99%) and located on the X chromosome (74%). Only one genomic region, containing the promoter of ZFP57, was hypomethylated in the pubertal group. ZFP57 is a transcriptional repressor required for both methylation and imprinting of multiple genomic loci. ZFP57 expression in the hypothalamus of female rhesus monkeys increased during peripubertal development, suggesting enhanced repression of downstream ZFP57 target genes. Fourteen other zinc finger (ZNF) genes were related to the hypermethylated DMRs at normal puberty. Analyzing the methylome changes associated with CPP, we demonstrated that the patients with CPP exhibited more hypermethylated CpG sites compared to both pre-pubertal (81%) and pubertal (89%) controls. Forty-eight ZNF genes were identified as having hypermethylated CpG sites in CPP.ConclusionMethylome profiling of girls at normal and precocious puberty revealed a widespread pattern of DNA hypermethylation, indicating that the pubertal process in humans is associated with specific changes in epigenetically driven regulatory control. Moreover, changes in methylation of several ZNF genes appear to be a distinct epigenetic modification underlying the initiation of human puberty.

Project description:BackgroundCentral precocious puberty (CPP) is a multifactorial and complex condition. Traditional studies focusing on a single indicator cannot always elucidate this panoramic condition but these may be revealed by using omics techniques.ObjectiveProteomics and metabolomics analysis of girls with CPP were compared to normal controls and the potential biomarkers and pathways involved were explored.MethodsSerum proteins and metabolites from normal girls and those with CPP were compared by LC-MS/MS. Multivariate and univariate statistical analysis were used to identify the differentially expressed proteins (DEPs) and differentially expressed metabolites (DEMs). Functional annotation and pathway enrichment analysis were performed by using GO and KEGG databases, and candidate markers were screened. Finally, bioinformatic analysis was used to integrate the results of proteomics and metabolomics to find the key differential proteins, metabolites and potential biomarkers of CPP.Results134 DEPs were identified in girls with CPP with 71 up- and 63 down-regulated, respectively. Up-regulated proteins were enriched mainly in the extracellular matrix, cell adhesion and cellular protein metabolic processes, platelet degranulation and skeletal system development. The down-regulated proteins were mainly enriched in the immune response. Candidate proteins including MMP9, TIMP1, SPP1, CDC42, POSTN, COL1A1, COL6A1, COL2A1 and BMP1, were found that may be related to pubertal development. 103 DEMs were identified, including 42 up-regulated and 61 down-regulated metabolites which were mainly enriched in lipid and taurine metabolic pathways. KGML network analysis showed that phosphocholine (16:1(9Z)/16:1(9Z)) was involved in arachidonic acid, glycerophospholipid, linoleic acid and α-linolenic acid metabolism and it may be used as a biomarker of CPP.ConclusionsOur study is the first to integrate proteomics and metabolomics to analyze the serum of girls with CPP and we found some key differential proteins and metabolites as well as a potential biomarker for this condition. Lipid metabolism pathways are involved and these may provide a key direction to further explore the molecular mechanisms and pathogenesis of CPP.

Dataset Information

Development of Prediction Models Using Machine Learning Algorithms for Girls with Suspected Central Precocious Puberty: Retrospective Study.

Publications

Development of Prediction Models Using Machine Learning Algorithms for Girls with Suspected Central Precocious Puberty: Retrospective Study.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets