title
  

PAMOGK: A pathway graph kernel based multi-omics clustering approach for discovering cancer patient subgroups

Warning The system is temporarily closed to updates for reporting purpose.

Tepeli, Yasin İlkağan and Ünal, Ali Burak and Akdemir, Mustafa Furkan and Taştan, Öznur (2020) PAMOGK: A pathway graph kernel based multi-omics clustering approach for discovering cancer patient subgroups. In: Satellite Workshop on Computational Cancer Biology (RECOMB-CCB) - Jointly with RECOMB 2020, Virtual

[img]
Preview
PDF - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
733Kb

Official URL: http://dx.doi.org/10.1101/834168

Abstract

Accurate classification of patients into homogeneous molecular subgroups is critical for the developmentof effective therapeutics and for deciphering what drives these different subtypes to cancer. However, the extensivemolecular heterogeneity observed among cancer patients presents a challenge. The availability of multi-omic datacatalogs for large cohorts of cancer patients provides multiple views into the molecular biology of the tumorswith unprecedented resolution. In this work, we develop PAMOGK, which integrates multi-omics patient data andincorporates the existing knowledge on biological pathways. PAMOGK is well suited to deal with the sparsity ofalterations in assessing patient similarities. We develop a novel graph kernel which we denote as smoothed shortestpath graph kernel, which evaluates patient similarities based on a single molecular alteration type in the contextof pathway. To corroborate multiple views of patients evaluated by hundreds of pathways and molecular alterationcombinations, PAMOGK uses multi-view kernel clustering. We apply PAMOGK to find subgroups of kidney renalclear cell carcinoma (KIRC) patients, which results in four clusters with significantly different survival times (p-value =7.4e-10). The patient subgroups also differ with respect to other clinical parameters such as tumor stage andgrade, and primary tumor and metastasis tumor spreads. When we compare PAMOGK to 8 other state-of-the-artexisting multi-omics clustering methods, PAMOGK consistently outperforms these in terms of its ability to partitionpatients into groups with different survival distributions. PAMOGK enables extracting the relative importance ofpathways and molecular data types. PAMOGK is available at github.com/tastanlab/pamogk

Item Type:Papers in Conference Proceedings
Additional Information:The paper is 10 page and peer-reviewed. Only the selected papers are invited for a talk. Our work was selected as a talk. There is no proceedings since we opt out of the option for the journal publication (it was under review in another journal). In place of a proceeding we were required to put the work on a preprint server. The work was under revision in a journal at the time. The journal submission and the conference submission versions are different. We improved the work in the journal.
Subjects:UNSPECIFIED
ID Code:40977
Deposited By:Öznur Taştan
Deposited On:27 Sep 2020 10:04
Last Modified:15 Oct 2020 10:15

Repository Staff Only: item control page