Community Assessment of the Predictability of Cancer Protein and Phosphoprotein Levels from Genomics and Transcriptomics

Mi Yang, Francesca Petralia, Zhi Li, Hongyang Li, Weiping Ma, Xiaoyu Song, Sunkyu Kim, Heewon Lee, Han Yu, Bora Lee, Seohui Bae, Eunji Heo, Jan Kaczmarczyk, Piotr Stępniak, Michał Warchoł, Thomas Yu, Anna P. Calinawan, Paul C. Boutros, Samuel H. Payne, Boris RevaTunde Aderinwale, Ebrahim Afyounian, Piyush Agrawal, Mehreen Ali, Alicia Amadoz, Francisco Azuaje, John Bachman, Sherry Bhalla, José Carbonell-Caballero, Priyanka Chakraborty, Kumardeep Chaudhary, Yonghwa Choi, Yoonjung Choi, Cankut Çubuk, Sandeep Kumar Dhanda, Joaquín Dopazo, Laura L. Elo, Ábel Fóthi, Olivier Gevaert, Kirsi Granberg, Russell Greiner, Marta R. Hidalgo, Vivek Jayaswal, Hwisang Jeon, Minji Jeon, Sunil V. Kalmady, Yasuhiro Kambara, Sang Yoon Kim, Arnaud Muller, Petr V. Nazarov, NCI-CPTAC-DREAM Consortium, Léon-Charles Tranchevent

Research output: Contribution to journalArticleResearchpeer-review

10 Citations (Scopus)

Abstract

A major manifestation of cancer is the alteration of protein measurements. However, proteins are harder and more expensive to measure than genes and transcripts. To address this problem, we crowdsourced it via the NCI-CPTAC DREAM proteogenomics challenge. We provided participants data to build models to predict protein and phosphorylation levels from genomic and transcriptomic data in cancer patients. We then asked participants to use such models to predict unseen (phospho)protein data from given genomic and transcriptomic data in other patients. This experiment allowed us to assess the predictive performance of the proposed methods in an unbiased and “double-blinded” manner. We found that ensemble methods perform better, and we identified which proteins and biological processes are easier or harder to predict. In general, performance was limited, suggesting that (phospho)proteomic cannot be replaced, at least yet, by genomic and transcriptomic profiling.

Original languageEnglish
Pages (from-to)186-195.e9
JournalCell Systems
Volume11
Issue number2
DOIs
Publication statusPublished - 26 Aug 2020

Keywords

  • cancer
  • crowdsourcing
  • genomics
  • machine learning
  • protein regulation
  • proteogenomics
  • proteomics

Fingerprint

Dive into the research topics of 'Community Assessment of the Predictability of Cancer Protein and Phosphoprotein Levels from Genomics and Transcriptomics'. Together they form a unique fingerprint.

Cite this