Explainable artificial intelligence model for identifying COVID-19 gene biomarkers

Abstract

Aim: COVID-19 has revealed the need for fast and reliable methods to assist clinicians in diagnosing the disease. This article presents a model that applies explainable artificial intelligence (XAI) methods based on machine learning techniques on COVID-19 metagenomic next-generation sequencing (mNGS) samples.

Methods: In the data set used in the study, there are 15,979 gene expressions of 234 patients with COVID-19 negative 141 (60.3%) and COVID-19 positive 93 (39.7%). The least absolute shrinkage and selection operator (LASSO) method was applied to select genes associated with COVID-19. Support Vector Machine - Synthetic Minority Oversampling Technique (SVM-SMOTE) method was used to handle the class imbalance problem. Logistics regression (LR), SVM, random forest (RF), and extreme gradient boosting (XGBoost) methods were constructed to predict COVID-19. An explainable approach based on local interpretable model-agnostic explanations (LIME) and SHAPley Additive exPlanations (SHAP) methods was applied to determine COVID-19- associated biomarker candidate genes and improve the final model's interpretability.

Results: For the diagnosis of COVID-19, the XGBoost (accuracy: 0.930) model outperformed the RF (accuracy: 0.912), SVM (accuracy: 0.877), and LR (accuracy: 0.912) models. As a result of the SHAP, the three most important genes associated with COVID-19 were IFI27, LGR6, and FAM83A. The results of LIME showed that especially the high level of IFI27 gene expression contributed to increasing the probability of positive class.

Conclusions: The proposed model (XGBoost) was able to predict COVID-19 successfully. The results show that machine learning combined with LIME and SHAP can explain the biomarker prediction for COVID-19 and provide clinicians with an intuitive understanding and interpretability of the impact of risk factors in the model.

Keywords: COVID-19; Explainable artificial intelligence; LIME; SHAP; XGBoost.

Fig. 1

Diagram of the proposed method combining explainability and classifier.

Fig. 2

The plot of results of ML models for COVID-19 (95% confidence interval (CI)).

Fig. 3

Interpretation of the XGBoost model. (A): Ranking the importance of the top 20 risk factors with stability and interpretation using the optimal model. (B): The order of importance of the first 20 variables according to the mean (|SHAP value|); the higher the SHAP value of a trait is given, the higher the probability that the patient will be COVID-19 positive.

Fig. 4

Variation of the five most important gene expression levels between groups.

Fig. 5

Radar plot of the five most important genes. - It is a COVID-19 patient appearing in orange on the Radar plot and a patient in the control group appearing in blue.

Fig. 6

Local interpretable model-agnostic explanations.