Multitask-Guided Self-Supervised Tabular Learning for Patient-Specific Survival Prediction

You Wu, Omid Bazgir, Yongju Lee, Tommaso Biancalani, James Lu, Ehsan Hajiramezanali
Proceedings of the 18th Machine Learning in Computational Biology meeting, PMLR 240:10-22, 2024.

Abstract

Survival prediction, central to the analysis of clinical trials, has the potential to be transformed by the availability of RNA-seq data as it reveals the underlying molecular and genetic mechanisms for disease and outcomes. However, the amount of RNA-seq samples available for understudied or rare diseases is often limited. To address this, leveraging data across different cancer types can be a viable solution, necessitating the application of self-supervised learning techniques. Yet, this wealth of data often comes in a tabular format without a known structure, hindering the development of a generally effective augmentation method for survival prediction. While traditional methods have been constrained by a one cancer-one model philosophy or have relied solely on a single modality, our approach, Guided-STab, on the contrary, offers a comprehensive approach through pretraining on all available RNA-seq data from various cancer types while guiding the representation by incorporating sparse clinical features as auxiliary tasks. With a multitask-guided self-supervised representation learning framework, we maximize the potential of vast unlabeled datasets from various cancer types, leading to genomic-driven survival predictions. These auxiliary clinical tasks then guide the learned representations to enhance critical survival factors. Extensive experiments reinforce the promise of our approach, as Guided-STab consistently outperforms established benchmarks on TCGA dataset.

Cite this Paper

BibTeX
@InProceedings{pmlr-v240-wu24a, title = {Multitask-Guided Self-Supervised Tabular Learning for Patient-Specific Survival Prediction}, author = {Wu, You and Bazgir, Omid and Lee, Yongju and Biancalani, Tommaso and Lu, James and Hajiramezanali, Ehsan}, booktitle = {Proceedings of the 18th Machine Learning in Computational Biology meeting}, pages = {10--22}, year = {2024}, editor = {Knowles, David A. and Mostafavi, Sara}, volume = {240}, series = {Proceedings of Machine Learning Research}, month = {30 Nov--01 Dec}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v240/wu24a/wu24a.pdf}, url = {https://proceedings.mlr.press/v240/wu24a.html}, abstract = {Survival prediction, central to the analysis of clinical trials, has the potential to be transformed by the availability of RNA-seq data as it reveals the underlying molecular and genetic mechanisms for disease and outcomes. However, the amount of RNA-seq samples available for understudied or rare diseases is often limited. To address this, leveraging data across different cancer types can be a viable solution, necessitating the application of self-supervised learning techniques. Yet, this wealth of data often comes in a tabular format without a known structure, hindering the development of a generally effective augmentation method for survival prediction. While traditional methods have been constrained by a one cancer-one model philosophy or have relied solely on a single modality, our approach, Guided-STab, on the contrary, offers a comprehensive approach through pretraining on all available RNA-seq data from various cancer types while guiding the representation by incorporating sparse clinical features as auxiliary tasks. With a multitask-guided self-supervised representation learning framework, we maximize the potential of vast unlabeled datasets from various cancer types, leading to genomic-driven survival predictions. These auxiliary clinical tasks then guide the learned representations to enhance critical survival factors. Extensive experiments reinforce the promise of our approach, as Guided-STab consistently outperforms established benchmarks on TCGA dataset.} }
Endnote
%0 Conference Paper %T Multitask-Guided Self-Supervised Tabular Learning for Patient-Specific Survival Prediction %A You Wu %A Omid Bazgir %A Yongju Lee %A Tommaso Biancalani %A James Lu %A Ehsan Hajiramezanali %B Proceedings of the 18th Machine Learning in Computational Biology meeting %C Proceedings of Machine Learning Research %D 2024 %E David A. Knowles %E Sara Mostafavi %F pmlr-v240-wu24a %I PMLR %P 10--22 %U https://proceedings.mlr.press/v240/wu24a.html %V 240 %X Survival prediction, central to the analysis of clinical trials, has the potential to be transformed by the availability of RNA-seq data as it reveals the underlying molecular and genetic mechanisms for disease and outcomes. However, the amount of RNA-seq samples available for understudied or rare diseases is often limited. To address this, leveraging data across different cancer types can be a viable solution, necessitating the application of self-supervised learning techniques. Yet, this wealth of data often comes in a tabular format without a known structure, hindering the development of a generally effective augmentation method for survival prediction. While traditional methods have been constrained by a one cancer-one model philosophy or have relied solely on a single modality, our approach, Guided-STab, on the contrary, offers a comprehensive approach through pretraining on all available RNA-seq data from various cancer types while guiding the representation by incorporating sparse clinical features as auxiliary tasks. With a multitask-guided self-supervised representation learning framework, we maximize the potential of vast unlabeled datasets from various cancer types, leading to genomic-driven survival predictions. These auxiliary clinical tasks then guide the learned representations to enhance critical survival factors. Extensive experiments reinforce the promise of our approach, as Guided-STab consistently outperforms established benchmarks on TCGA dataset.
APA
Wu, Y., Bazgir, O., Lee, Y., Biancalani, T., Lu, J. & Hajiramezanali, E.. (2024). Multitask-Guided Self-Supervised Tabular Learning for Patient-Specific Survival Prediction. Proceedings of the 18th Machine Learning in Computational Biology meeting, in Proceedings of Machine Learning Research 240:10-22 Available from https://proceedings.mlr.press/v240/wu24a.html.

Related Material