NCI Data Jamboree (Project Abstract Submission): Submission #19

Presenter Information
---------------------
First Name: Hongyi
Middle Initial: {Empty}
Last Name: Liu
Degree(s): Ph.D.
Position/Title/Career Status: {Empty}
Organization: Johns Hopkins University
Organization Address:
Baltimore

Email: hliu173@jh.edu

Additional Authors
------------------
List of Additional Authors:
- First Name: Yingwei
  Last Name: Hu
  Affiliation: Johns Hopkins University
- First Name: Mamie
  Last Name: Lih
  Affiliation: Johns Hopkins University
- First Name: Effram
  Last Name: Wei
  Affiliation: Johns Hopkins University
- First Name: Liyuan
  Last Name: Jiao
  Affiliation: Johns Hopkins University
- First Name: Lijun
  Last Name: Chen
  Affiliation: Johns Hopkins University
- First Name: Yuefan
  Last Name: Wang
  Affiliation: Johns Hopkins University
- First Name: Xiangning
  Last Name: Li
  Affiliation: Johns Hopkins University
- First Name: Zhenyu
  Last Name: Sun
  Affiliation: Johns Hopkins University
- First Name: Yuanyu
  Last Name: Huang
  Affiliation: Johns Hopkins University
- First Name: Yuanwei
  Last Name: Xu
  Affiliation: Johns Hopkins University
- First Name: Hui
  Last Name: Zhang
  Affiliation: Johns Hopkins University


Abstract Information
--------------------
Abstract Category: Employing statistical, computational, and informatics tools, algorithms, and methods to integrate or analyze data
Abstract Keywords: Prostate cancer, proteomics, glycoproteomics, phosphoproteomics, crosstalk
Abstract Title: Unlocks the Biological Insights into Aggressive Prostate Cancer Using Multi-omic Approach
Abstract:
Scientific Questions: This project integrates large-scale multi-omics to address: (1) how multi-layer genomics, proteomics (TMT/DIA), and post-translational modifications (PTMs) correlate globally; (2) how to digitally deconvolve tumor microenvironment (TME) cell fractions; and (3) how to map the causal directional regulatory cascades between cell-surface glycosylation and intracellular phosphorylation to identify therapeutic targets.
Community Significance: Prostate cancer has profound molecular heterogeneity. Providing a reproducible, open-source computational framework for integrating mass-spectrometry-based proteomics and dual-PTM networks allows the broader cancer research community to uncover targetable biological pathways and patient subclusters obscured in genomic-only studies.
Datasets & Tools: We leverage clinical cohorts (244 samples) with comprehensive transcriptomics, global proteomics (TMT/DIA), and deeply enriched PTMs (phospho, intact glycopeptide, ubiquitin, p-Tyr, acetyl). Key informatics tools include xCell/CIBERSORTx for cellular deconvolution, Random Forest and Lasso-logistic regression for tumor grading classifiers (NAT vs. Low/High Gleason grades), human interactome mapping for protein-protein interaction (PPI) topology, and Bayesian modeling for causal dual-PTM network inference.