A meta-learning approach for genomic survival analysis

Abstract

RNA sequencing has emerged as a promising approach in cancer prognosis as sequencing data becomes more easily and affordably accessible. However, it remains challenging to build good predictive models especially when the sample size is limited and the number of features is high, which is a common situation in biomedical settings. To address these limitations, we propose a meta-learning framework based on neural networks for survival analysis and evaluate it in a genomic cancer research setting. We demonstrate that, compared to regular transfer-learning, meta-learning is a significantly more effective paradigm to leverage high-dimensional data that is relevant but not directly related to the problem of interest. Specifically, meta-learning explicitly constructs a model, from abundant data of relevant tasks, to learn a new task with few samples effectively. For the application of predicting cancer survival outcome, we also show that the meta-learning framework with a few samples is able to achieve competitive performance with learning from scratch with a significantly larger number of samples. Finally, we demonstrate that the meta-learning model implicitly prioritizes genes based on their contribution to survival prediction and allows us to identify important pathways in cancer.

A Few-Shot Learning Approach Assists in the Prognosis Prediction of Magnetic Resonance-Guided Focused Ultrasound for the Local Control of Bone Metastatic Lesions

“…In order to overcome the limitation of the few patients, we suggest a novel application of few-shot learning algorithm for predict the prognosis after MRgFUS palliative treatment. The few-shot learning approach was specified to extract eigenvalues and classify hidden subsets [ 31 , 32 , 34 , 39 ]. The algorithm embedded a Bayesian framework that minimized training samples by prior parameter selections.…”

Artificial intelligence for clinical oncology

“…This classifier has been subsequently validated in several external settings, and is now undergoing investigation in randomized controlled trials (NCT04513717,NCT02783950). Deep-learning strategies have been explored to integrate multi-omic data sources into riskstratification models utilizing combinations of diagnostic imaging (Kann et al, 2020b), EHR data (Beg et al, 2017;Manz et al, 2020), and genomic information (Qiu et al, 2020). Furthermore, there is the potential for deep learning to better risk-stratify (She et al, 2020).…”

Artificial Intelligence in Bulk and Single-Cell RNA-Sequencing Data to Foster Precision Oncology

“…In cancer transcriptomics, ML and DL models have been applied to classify different cancer subtypes and cell populations [ 17 , 18 , 19 , 20 ], characterize tumor immune microenvironment [ 21 , 22 , 23 , 24 , 25 ], discover new prognostic biomarkers [ 26 , 27 , 28 ], assess and predict disease recurrence and patient survival [ 29 , 30 , 31 , 32 ], identify new putative actionable vulnerabilities [ 33 , 34 ], and predict tumor antigen immunogenicity [ 35 ] ( Figure 2 ).…”

“…For this reason, ML extensions of the Cox-PH model employing random forest have been developed [ 117 ]. Recently, NN approaches have been shown to outperform classical survival methods [ 29 , 30 , 31 , 32 ]. These algorithms exploit feature selection through NNs to obtain a subset of surrogate prognostic features.…”

“…Transfer learning is a ML technique by which a model trained on one setting is exploited on another related setting [ 118 ]. Transfer learning has been employed to assess patient survival in pan-cancer RNA-seq datasets [ 29 ]. The model showed a higher prognostic performance than competing methods and exploiting risk score backpropagation [ 119 ] allowed to assess the biological pathways that impact on patient’s survival outcome in the tested cancer types.…”