Survival analysis plays a critical role in cancer research, offering insights into patient outcomes and informing personalized treatments. Advances in sequencing and profiling technologies have unlocked the potential of multi-omics data to improve prognostic models. However, the high dimensionality of multi-omics data often complicates analysis and limits clinical application. Additionally, most studies rely on the traditional Cox proportional hazards model, with limited exploration of alternative survival algorithms or feature selection methods. Few frameworks integrate optimal features across multiple omics modalities, leaving a gap in harnessing the full potential of multi-omics data. To address these challenges, we developed PRISM (PRognostic marker Identification and Survival Modelling through Multi-omics Integration), a comprehensive framework for improving survival predictions and identifying key prognostic markers. PRISM systematically evaluates feature selection methods and survival models, employing a voting-based pipeline to identify robust features from single-omics data. These features are integrated using feature-level fusion to determine the best modality combinations for survival prediction. Applied to The Cancer Genome Atlas (TCGA) data for Breast Invasive Carcinoma (BRCA), Cervical Squamous Cell Carcinoma and Endocervical Adenocarcinoma (CESC), and Uterine Corpus Endometrial Carcinoma (UCEC), PRISM demonstrated that integrating DNA methylation, miRNA, and copy number variation data significantly outperforms single-omics approaches (BRCA: C-index 0.774; CESC: 0.835; UCEC: 0.766). Pan-cancer analysis further demonstrated the biological relevance of these multi-modal signatures, revealing shared oncogenic pathways and potential therapeutic targets. By reducing data dimensionality while preserving predictive power, PRISM offers a scalable and general-purpose solution for integrating multi-omics data, advancing cancer research and enabling precision medicine.

Highlights

Introducing PRISM, a robust framework that enhances survival predictions and identifies key prognostic markers through multi-omics integration.

Utilizing a voting-based pipeline, PRISM selects optimal single-omics features and integrates them via feature-level fusion to identify the best modality combinations.

Highlighting Survival Random Forest as an effective model for high-dimensional data, outperforming other survival algorithms.

Demonstrating superior survival predictions with integrated DNA methylation, miRNA, and copy number variation data in Breast Invasive Carcinoma (C-index 0.774), Cervical Squamous Cell Carcinoma and Endocervical Adenocarcinoma (CESC, 0.835), and Uterine Corpus Endometrial Carcinoma (UCEC, 0.766).

Revealing common oncogenic pathways and therapeutic targets through pan-cancer analysis validated by Kaplan-Meier analysis of multi-modal signatures.

The authors have declared no competing interest.

This study did not receive any funding

I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.

Yes

The details of the IRB/oversight body that provided approval or exemption for the research described are given below:

TCGA Data from the GDC Portal

I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals.

Yes

I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).

Yes

I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable.

Yes