Machine Learning to Improve Accuracy of Transcutaneous Bilirubinometry.

INTRODUCTION: This study aimed to develop models for predicting total serum bilirubin by correcting errors of transcutaneous bilirubin using machine learning based on neonatal biomarkers that could affect spectrophotometric measurements of tissue bilirubin. METHODS: This retrospective study included infants born at our hospital (≥36 weeks old, ≥2,000 g) between January 2020 and December 2022. Infants without a phototherapy history were included. Robust linear regression, gradient boosting tree, and neural networks were used for machine learning models. A neural network, inspired by the structure of the human brain, was designed comprising three layers: input, intermediate, and output. RESULTS: Totally, 683 infants were included. The mean (minimum-maximum) gestational age, birth weight, participant age, total serum bilirubin, and transcutaneous bilirubin were 39.0 (36.0-42.0) weeks, 3,004 (2,004-4,484) g, 2.8 (1-6) days of age, 8.50 (2.67-18.12) mg/dL, and 7.8 (1.1-18.1) mg/dL, respectively. The neural network model had a root mean square error of 1.03 mg/dL and a mean absolute error of 0.80 mg/dL in cross-validation data. These values were 0.37 mg/dL and 0.28 mg/dL, smaller compared to transcutaneous bilirubin, respectively. The 95% limit of agreement between the neural network estimation and total serum bilirubin was -2.01 to 2.01 mg/dL. Unnecessary blood draws could be reduced by up to 78%. CONCLUSION: Using machine learning with transcutaneous bilirubin, total serum bilirubin estimation error was reduced by 25%. This integration could increase accuracy, lessen infant discomfort, and simplify procedures, offering a smart alternative to blood draws by accurately estimating phototherapy thresholds.

The regulation mechanisms of AhR by molecular chaperone complex.

The AhR, so called the dioxin receptor, is a member of the nuclear receptor superfamily. The ligand-free AhR forms a cytosolic protein complex with the molecular chaperone HSP90, co-chaperone p23, and XAP2 in the cytoplasm. Following ligand binding like 2, 3, 7, 8-tetrachlorodibenzo-p-dioxin (TCDD), the AhR translocates into the nucleus. Although it has been reported that HSP90 regulates the translocation of the AhR to the nucleus, the precise activation mechanisms of the AhR have not yet been fully understood. AhR consists of the N-terminal bHLH domain containing NLS and NES, the middle PAS domain and the C-terminal transactivation domain. The PAS domain is familiar as a ligand and HSP90 binding domain. In this study, we focused on the bHLH domain that was thought to be a HSP90 binding domain. We investigated the binding properties of bHLH to HSP90. We analyzed the direct interaction of bHLH with HSP90, p23 and XAP2 using purified proteins. We found that not only the PAS domain but also the bHLH domain bound to HSP90. The bHLH domain forms complex with HSP90, p23 and XAP2. We also determined the bHLH binding domain was HSP90 N-domain. The bHLH domain makes a complex with HSP90, p23 and XAP2 via the HSP90 N-domain. Although the NLS is closed in the absence of a ligand, the structure of AhR will be changed in the presence of a ligand, which leads to NLS open, result in the nuclear translocation of AhR.

The activation mechanism of the aryl hydrocarbon receptor (AhR) by molecular chaperone HSP90.

The aryl hydrocarbon receptor is a member of the nuclear receptor superfamily that associates with the molecular chaperone HSP90 in the cytoplasm. The activation mechanism of the AhR is not yet fully understood. It has been proposed that after binding of ligands such as 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), 3methylcholanthrene (3-MC), or ß-naphthoflavone (ß-NF), the AhR dissociates from HSP90 and translocates to the nucleus. It has also been hypothesized that the AhR translocates to the nucleus and forms a complex with HSP90 and other co-chaperones. There are a few reports about the direct association or dissociation of AhR and HSP90 due to difficulties in purifying AhR. We constructed and purified the PAS domain from AhR. Binding of the AhR-PAS domain to ß-NF affinity resin suggested that it possesses ligand-binding affinity. We demonstrated that the AhR-PAS domain binds to HSP90 and the association is not affected by ligand binding. The ligand 17-DMAG inhibited binding of HSP90 to GST-PAS. In an immunoprecipitation assay, HSP90 was co-immunoprecipitated with AhR both in the presence or absence of ligand. Endogenous AhR decreased in the cytoplasm and increased in the nucleus of HeLa cells 15 min after treatment with ligand. These results suggested that the ligand-bound AhR is translocated to nucleus while in complex with HSP90. We used an in situ proximity ligation assay to confirm whether AhR was translocated to the nucleus alone or together with HSP90. HSP90 was co-localized with AhR after the nuclear translocation. It has been suggested that the ligand-bound AhR was translocated to the nucleus with HSP90. Activated AhR acts as a transcription factor, as shown by the transcription induction of the gene CYP1A1 8 h after treatment with ß-NF.