Exploring combinations of dimensionality reduction, transfer learning, and regularization methods for predicting binary phenotypes with transcriptomic data

Abstract

Abstract Background Numerous transcriptomic-based models have been developed to predict or understand the fundamental mechanisms driving biological phenotypes. However, few models have successfully transitioned into clinical practice due to challenges associated with generalizability and interpretability. To address these issues, researchers have turned to dimensionality reduction methods and have begun implementing transfer learning approaches. Methods In this study, we aimed to evaluate the effectiveness of these strategies by exploring the optimal combination of dimensionality reduction methods (with and without transfer learning), and regularization techniques in predictive modeling. We employed four dimensionality reduction methods, namely, Principal Component Analysis (PCA), Consensus Independent Component Analysis (c-ICA), Autoencoder (AE), and Adversarial Variational Autoencoder (AVAE). Additionally, we applied a transfer learning approach by training the AE, AVEA, and c-ICA models on approximately 140,000 transcriptomic profiles. To assess the performance of the different combinations, we used a cross-validation setup encapsulated within a permutation testing framework, analyzing 30 different transcriptomic datasets with binary phenotypes. Furthermore, we included datasets with small sample sizes and phenotypes of varying degrees of predictability, and we employed independent datasets for validation. Results Our findings revealed that regularized models without dimensionality reduction achieved the highest predictive performance, challenging the necessity of dimensionality reduction when the primary goal is to achieve optimal predictive performance. However, models using AE and c-ICA with transfer learning for dimensionality reduction showed comparable performance, with enhanced interpretability and robustness of predictors, compared to models using non-dimensionality-reduced data. Conclusion These findings offer valuable insights into the optimal combination of strategies for enhancing the predictive performance, interpretability, and generalizability of transcriptomic-based models.