Predicting Metabolite–Disease Associations Based on LightGBM Model

Zhang, Cheng; Lei, Xiujuan

doi:10.3389/fgene.2021.660275

Cited by 9 publications

(4 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In order to obtain better performance of the GCHIRFLDA model, we compared RF classifier with several classical classifiers, including extreme gradient boosting (Xgboost) ( Chen and Guestrin, 2016 ), C50 ( Kuhn, 2013 ), Gradient Boosting Decision Tree (GBDT) ( Ye et al, 2009 ), SVM ( Lan et al, 2017 ) and LightGBM ( Zhang et al, 2021 ). In this work, we used the average AUC, AUPR, Recall, F1-score and Accuracy based on five-fold cross-validation as evaluation criterion for the six classifiers.…”

Section: Resultsmentioning

confidence: 99%

Geometric complement heterogeneous information and random forest for predicting lncRNA-disease associations

Yao

Zhang²,

Zhan³

et al. 2022

Front. Genet.

View full text Add to dashboard Cite

More and more evidences have showed that the unnatural expression of long non-coding RNA (lncRNA) is relevant to varieties of human diseases. Therefore, accurate identification of disease-related lncRNAs can help to understand lncRNA expression at the molecular level and to explore more effective treatments for diseases. Plenty of lncRNA-disease association prediction models have been raised but it is still a challenge to recognize unknown lncRNA-disease associations. In this work, we have proposed a computational model for predicting lncRNA-disease associations based on geometric complement heterogeneous information and random forest. Firstly, geometric complement heterogeneous information was used to integrate lncRNA-miRNA interactions and miRNA-disease associations verified by experiments. Secondly, lncRNA and disease features consisted of their respective similarity coefficients were fused into input feature space. Thirdly, an autoencoder was adopted to project raw high-dimensional features into low-dimension space to learn representation for lncRNAs and diseases. Finally, the low-dimensional lncRNA and disease features were fused into input feature space to train a random forest classifier for lncRNA-disease association prediction. Under five-fold cross-validation, the AUC (area under the receiver operating characteristic curve) is 0.9897 and the AUPR (area under the precision-recall curve) is 0.7040, indicating that the performance of our model is better than several state-of-the-art lncRNA-disease association prediction models. In addition, case studies on colon and stomach cancer indicate that our model has a good ability to predict disease-related lncRNAs.

show abstract

Section: Resultsmentioning

confidence: 99%

Geometric complement heterogeneous information and random forest for predicting lncRNA-disease associations

Yao

Zhang²,

Zhan³

et al. 2022

Front. Genet.

View full text Add to dashboard Cite

show abstract

“…(A) LightGBM is a gradient-boosting framework based on a decision tree (DT). It uses a node segmentation strategy based on leaves, seeks the leaf with the largest gain among all the current leaves, and finally generates a boosted tree ( 45 , 46 ). The LightGBM algorithm is based on the selection of partition points based on the histogram algorithm and reduces the number of samples and features required in the training and learning processes through two methods, namely, gradient-based one-side sampling (GOSS) and exclusive feature bundling (EFB), to maintain high learning performance and reduce the resource occupation in terms of time and space in the training process ( 47 , 48 ).…”

Section: Methodsmentioning

confidence: 99%

OEDL: an optimized ensemble deep learning method for the prediction of acute ischemic stroke prognoses using union features

Chen

et al. 2023

Front. Neurol.

View full text Add to dashboard Cite

BackgroundEarly stroke prognosis assessments are critical for decision-making regarding therapeutic intervention. We introduced the concepts of data combination, method integration, and algorithm parallelization, aiming to build an integrated deep learning model based on a combination of clinical and radiomics features and analyze its application value in prognosis prediction.MethodsThe research steps in this study include data source and feature extraction, data processing and feature fusion, model building and optimization, model training, and so on. Using data from 441 stroke patients, clinical and radiomics features were extracted, and feature selection was performed. Clinical, radiomics, and combined features were included to construct predictive models. We applied the concept of deep integration to the joint analysis of multiple deep learning methods, used a metaheuristic algorithm to improve the parameter search efficiency, and finally, developed an acute ischemic stroke (AIS) prognosis prediction method, namely, the optimized ensemble of deep learning (OEDL) method.ResultsAmong the clinical features, 17 features passed the correlation check. Among the radiomics features, 19 features were selected. In the comparison of the prediction performance of each method, the OEDL method based on the concept of ensemble optimization had the best classification performance. In the comparison to the predictive performance of each feature, the inclusion of the combined features resulted in better classification performance than that of the clinical and radiomics features. In the comparison to the prediction performance of each balanced method, SMOTEENN, which is based on a hybrid sampling method, achieved the best classification performance than that of the unbalanced, oversampled, and undersampled methods. The OEDL method with combined features and mixed sampling achieved the best classification performance, with 97.89, 95.74, 94.75, 94.03, and 94.35% for Macro-AUC, ACC, Macro-R, Macro-P, and Macro-F1, respectively, and achieved advanced performance in comparison with that of methods in previous studies.ConclusionThe OEDL approach proposed herein could effectively achieve improved stroke prognosis prediction performance, the effect of using combined data modeling was significantly better than that of single clinical or radiomics feature models, and the proposed method had a better intervention guidance value. Our approach is beneficial for optimizing the early clinical intervention process and providing the necessary clinical decision support for personalized treatment.

show abstract

“…( 25 ). Currently, this framework has been relatively widely used in the field of medical data processing ( 26 – 28 ), but it has not been attempted in the HCC recurrence prediction task.…”

Section: Methodsmentioning

confidence: 99%

Ensemble learning based on efficient features combination can predict the outcome of recurrence-free survival in patients with hepatocellular carcinoma within three years after surgery

Wang

Zhu

et al. 2022

Front. Oncol.

View full text Add to dashboard Cite

Preoperative prediction of recurrence outcome in hepatocellular carcinoma (HCC) facilitates physicians’ clinical decision-making. Preoperative imaging and related clinical baseline data of patients are valuable for evaluating prognosis. With the widespread application of machine learning techniques, the present study proposed the ensemble learning method based on efficient feature representations to predict recurrence outcomes within three years after surgery. Radiomics features during arterial phase (AP) and clinical data were selected for training the ensemble models. In order to improve the efficiency of the process, the lesion area was automatically segmented by 3D U-Net. It was found that the mIoU of the segmentation model was 0.8874, and the Light Gradient Boosting Machine (LightGBM) was the most superior, with an average accuracy of 0.7600, a recall of 0.7673, a F1 score of 0.7553, and an AUC of 0.8338 when inputting radiomics features during AP and clinical baseline indicators. Studies have shown that the proposed strategy can relatively accurately predict the recurrence outcome within three years, which is helpful for physicians to evaluate individual patients before surgery.

show abstract

Predicting Metabolite–Disease Associations Based on LightGBM Model

Cited by 9 publications

References 22 publications

Geometric complement heterogeneous information and random forest for predicting lncRNA-disease associations

Geometric complement heterogeneous information and random forest for predicting lncRNA-disease associations

OEDL: an optimized ensemble deep learning method for the prediction of acute ischemic stroke prognoses using union features

Ensemble learning based on efficient features combination can predict the outcome of recurrence-free survival in patients with hepatocellular carcinoma within three years after surgery

Contact Info

Product

Resources

About