A Prediction Model for Tacrolimus Daily Dose in Kidney Transplant Recipients With Machine Learning and Deep Learning Techniques
Tacrolimus is a major immunosuppressor against post-transplant rejection in kidney transplant recipients. However, the narrow therapeutic index of tacrolimus and considerable variability among individuals are challenges for therapeutic outcomes. The aim of this study was to compare different machine...
| Main Authors: | , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2022-05-01
|
| Series: | Frontiers in Medicine |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fmed.2022.813117/full |
| _version_ | 1818240190560337920 |
|---|---|
| author | Qiwen Zhang Qiwen Zhang Xueke Tian Xueke Tian Guang Chen Guang Chen Ze Yu Xiaojian Zhang Xiaojian Zhang Jingli Lu Jingli Lu Jinyuan Zhang Peile Wang Peile Wang Xin Hao Yining Huang Zeyuan Wang Fei Gao Jing Yang Jing Yang |
| author_facet | Qiwen Zhang Qiwen Zhang Xueke Tian Xueke Tian Guang Chen Guang Chen Ze Yu Xiaojian Zhang Xiaojian Zhang Jingli Lu Jingli Lu Jinyuan Zhang Peile Wang Peile Wang Xin Hao Yining Huang Zeyuan Wang Fei Gao Jing Yang Jing Yang |
| author_sort | Qiwen Zhang |
| collection | DOAJ |
| description | Tacrolimus is a major immunosuppressor against post-transplant rejection in kidney transplant recipients. However, the narrow therapeutic index of tacrolimus and considerable variability among individuals are challenges for therapeutic outcomes. The aim of this study was to compare different machine learning and deep learning algorithms and establish individualized dose prediction models by using the best performing algorithm. Therefore, among the 10 commonly used algorithms we compared, the TabNet algorithm outperformed other algorithms with the highest R2 (0.824), the lowest prediction error [mean absolute error (MAE) 0.468, mean square error (MSE) 0.558, and root mean square error (RMSE) 0.745], and good performance of overestimated (5.29%) or underestimated dose percentage (8.52%). In the final prediction model, the last tacrolimus daily dose, the last tacrolimus therapeutic drug monitoring value, time after transplantation, hematocrit, serum creatinine, aspartate aminotransferase, weight, CYP3A5, body mass index, and uric acid were the most influential variables on tacrolimus daily dose. Our study provides a reference for the application of deep learning technique in tacrolimus dose estimation, and the TabNet model with desirable predictive performance is expected to be expanded and applied in future clinical practice. |
| first_indexed | 2024-12-12T13:09:30Z |
| format | Article |
| id | doaj.art-43bbe06659384a4585a3c4989e9573f9 |
| institution | Directory Open Access Journal |
| issn | 2296-858X |
| language | English |
| last_indexed | 2024-12-12T13:09:30Z |
| publishDate | 2022-05-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Medicine |
| spelling | doaj.art-43bbe06659384a4585a3c4989e9573f92022-12-22T00:23:33ZengFrontiers Media S.A.Frontiers in Medicine2296-858X2022-05-01910.3389/fmed.2022.813117813117A Prediction Model for Tacrolimus Daily Dose in Kidney Transplant Recipients With Machine Learning and Deep Learning TechniquesQiwen Zhang0Qiwen Zhang1Xueke Tian2Xueke Tian3Guang Chen4Guang Chen5Ze Yu6Xiaojian Zhang7Xiaojian Zhang8Jingli Lu9Jingli Lu10Jinyuan Zhang11Peile Wang12Peile Wang13Xin Hao14Yining Huang15Zeyuan Wang16Fei Gao17Jing Yang18Jing Yang19Department of Pharmacy, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, ChinaHenan Key Laboratory of Precision Clinical Pharmacy, Zhengzhou University, Zhengzhou, ChinaDepartment of Pharmacy, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, ChinaHenan Key Laboratory of Precision Clinical Pharmacy, Zhengzhou University, Zhengzhou, ChinaDepartment of Pharmacy, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, ChinaHenan Key Laboratory of Precision Clinical Pharmacy, Zhengzhou University, Zhengzhou, ChinaBeijing Medicinovo Technology Co. Ltd, Beijing, ChinaDepartment of Pharmacy, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, ChinaHenan Key Laboratory of Precision Clinical Pharmacy, Zhengzhou University, Zhengzhou, ChinaDepartment of Pharmacy, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, ChinaHenan Key Laboratory of Precision Clinical Pharmacy, Zhengzhou University, Zhengzhou, ChinaBeijing Medicinovo Technology Co. Ltd, Beijing, ChinaDepartment of Pharmacy, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, ChinaHenan Key Laboratory of Precision Clinical Pharmacy, Zhengzhou University, Zhengzhou, ChinaDalian Medicinovo Technology Co. Ltd, Dalian, ChinaMcCormick School of Engineering, Northwestern University, Evanston, IL, United StatesBeijing Medicinovo Technology Co. Ltd, Beijing, ChinaBeijing Medicinovo Technology Co. Ltd, Beijing, ChinaDepartment of Pharmacy, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, ChinaHenan Key Laboratory of Precision Clinical Pharmacy, Zhengzhou University, Zhengzhou, ChinaTacrolimus is a major immunosuppressor against post-transplant rejection in kidney transplant recipients. However, the narrow therapeutic index of tacrolimus and considerable variability among individuals are challenges for therapeutic outcomes. The aim of this study was to compare different machine learning and deep learning algorithms and establish individualized dose prediction models by using the best performing algorithm. Therefore, among the 10 commonly used algorithms we compared, the TabNet algorithm outperformed other algorithms with the highest R2 (0.824), the lowest prediction error [mean absolute error (MAE) 0.468, mean square error (MSE) 0.558, and root mean square error (RMSE) 0.745], and good performance of overestimated (5.29%) or underestimated dose percentage (8.52%). In the final prediction model, the last tacrolimus daily dose, the last tacrolimus therapeutic drug monitoring value, time after transplantation, hematocrit, serum creatinine, aspartate aminotransferase, weight, CYP3A5, body mass index, and uric acid were the most influential variables on tacrolimus daily dose. Our study provides a reference for the application of deep learning technique in tacrolimus dose estimation, and the TabNet model with desirable predictive performance is expected to be expanded and applied in future clinical practice.https://www.frontiersin.org/articles/10.3389/fmed.2022.813117/fullprediction modeltacrolimusdaily dosekidney transplantmachine learninggenetic polymorphism |
| spellingShingle | Qiwen Zhang Qiwen Zhang Xueke Tian Xueke Tian Guang Chen Guang Chen Ze Yu Xiaojian Zhang Xiaojian Zhang Jingli Lu Jingli Lu Jinyuan Zhang Peile Wang Peile Wang Xin Hao Yining Huang Zeyuan Wang Fei Gao Jing Yang Jing Yang A Prediction Model for Tacrolimus Daily Dose in Kidney Transplant Recipients With Machine Learning and Deep Learning Techniques Frontiers in Medicine prediction model tacrolimus daily dose kidney transplant machine learning genetic polymorphism |
| title | A Prediction Model for Tacrolimus Daily Dose in Kidney Transplant Recipients With Machine Learning and Deep Learning Techniques |
| title_full | A Prediction Model for Tacrolimus Daily Dose in Kidney Transplant Recipients With Machine Learning and Deep Learning Techniques |
| title_fullStr | A Prediction Model for Tacrolimus Daily Dose in Kidney Transplant Recipients With Machine Learning and Deep Learning Techniques |
| title_full_unstemmed | A Prediction Model for Tacrolimus Daily Dose in Kidney Transplant Recipients With Machine Learning and Deep Learning Techniques |
| title_short | A Prediction Model for Tacrolimus Daily Dose in Kidney Transplant Recipients With Machine Learning and Deep Learning Techniques |
| title_sort | prediction model for tacrolimus daily dose in kidney transplant recipients with machine learning and deep learning techniques |
| topic | prediction model tacrolimus daily dose kidney transplant machine learning genetic polymorphism |
| url | https://www.frontiersin.org/articles/10.3389/fmed.2022.813117/full |
| work_keys_str_mv | AT qiwenzhang apredictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT qiwenzhang apredictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT xueketian apredictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT xueketian apredictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT guangchen apredictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT guangchen apredictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT zeyu apredictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT xiaojianzhang apredictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT xiaojianzhang apredictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT jinglilu apredictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT jinglilu apredictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT jinyuanzhang apredictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT peilewang apredictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT peilewang apredictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT xinhao apredictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT yininghuang apredictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT zeyuanwang apredictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT feigao apredictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT jingyang apredictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT jingyang apredictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT qiwenzhang predictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT qiwenzhang predictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT xueketian predictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT xueketian predictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT guangchen predictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT guangchen predictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT zeyu predictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT xiaojianzhang predictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT xiaojianzhang predictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT jinglilu predictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT jinglilu predictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT jinyuanzhang predictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT peilewang predictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT peilewang predictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT xinhao predictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT yininghuang predictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT zeyuanwang predictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT feigao predictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT jingyang predictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques AT jingyang predictionmodelfortacrolimusdailydoseinkidneytransplantrecipientswithmachinelearninganddeeplearningtechniques |