Interpretable meta-learning of multi-omics data for survival analysis and pathway enrichment

Abstract

Motivation Despite the success of recent machine learning algorithms’ applications to survival analysis, their black-box nature hinders interpretability, which is arguably the most important aspect. Similarly, multi-omics data integration for survival analysis is often constrained by the underlying relationships and correlations that are rarely well understood. The goal of this work is to alleviate the interpretability problem in machine learning approaches for survival analysis and also demonstrate how multi-omics data integration improves survival analysis and pathway enrichment. We use meta-learning, a machine learning algorithm that is trained on a variety of related datasets and allows quick adaptations to new tasks, to perform survival analysis and pathway enrichment on pan-cancer datasets. In recent machine learning research, meta-learning has been effectively used for knowledge transfer among multiple related datasets. Results We use meta-learning with cox hazard loss to show that the integration of TCGA pan-cancer data increases the performance of survival analysis. We also apply advanced model interpretability method called DeepLIFT (Deep Learning Important FeaTures) to show different sets of enriched pathways for multi-omics and transcriptomics data. Our results show that multi-omics cancer survival analysis enhances performance compared with using transcriptomics or clinical data alone. Additionally, we show a correlation between variable importance assignment from DeepLIFT and gene co-enrichment, suggesting that genes with higher and similar contribution scores are more likely to be enriched together in the same enrichment sets. Availability https://github.com/berkuva/TCGA-omics-integration Supplementary information Supplementary data are available at Bioinformatics online.

Deep learning for survival analysis: a review

“…It is worth noting that post-hoc methods from the field of Interpretable Machine Learning and explainable AI, such as Permutation Feature Importance (Breiman 2001) Several survival-specific adaptations of the abovementioned post-hoc interpretability methods have already been developed; for instance, SurvLIME (Kovalev et al 2020), SurvSHAP(t) (Krzyziński et al 2022), and SurvNAM (Utkin et al 2022) are based on LIME, SHAP, and NAMs, respectively. Cho et al (2023) use meta-learning and the DeepLIFT (Shrikumar et al 2017) method to make the integration of multi-omics data in SA more interpretable.…”

A review of multi-omics data integration through deep learning approaches for disease diagnosis, prognosis, and treatment

“…The tools incorporate concepts such as loss functions meant for obtaining puissant feature learning and prediction ability. For instance, a meta-learning deep learning method termed DeepLIFT was recently proposed that implements cox hazard loss to improve performance, intelligibility and interpretability of the model ( Cho et al, 2023 ). Meta-learning, a learning-to-learn method, based on back propagation and cox hazard loss trained on transcriptomics, proteomics, and clinical datasets showed better performance than direct and transfer learning-based models.…”

IBPGNET: lung adenocarcinoma recurrence prediction based on neural network interpretability

“…Therefore, balancing the accuracy and interpretability of the deep learning model is very important in biomedical research. Many interpretable methods, such as LIME [ 14 ], RISE [ 15 ], Grad-CAM [ 16 ] and DeepLIFT [ 17 ], have been spawned to improve the interpretability of deep learning models. Besides, some methods inspired by biological systems have been successfully applied to cancer classification prediction by constructing neural networks with biological information [ 18–21 ].…”