Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Linkomics Kb topic
No spam. Unsubscribe anytime.
Researchers demonstrate LinkOmics KB for harmonized proteogenomic queries across CPTAC cohorts
Summary
Dr. Bing Zang demonstrated LinkOmics KB, a gene/protein/mutation/phenotype web resource built from harmonized CPTAC data across >1,000 tumors and 10 cancer types, and presented case studies showing how protein‑level insights can differ from RNA measures and highlight testable biological hypotheses.
Dr. Bing Zang of Baylor College of Medicine described LinkOmics KB, a web resource built on harmonized CPTAC proteogenomic data, and showed examples of how integrated protein, RNA and mutation data can generate testable biological insights. He said the resource includes roughly 40,000 dedicated pages (gene, protein isoform, mutation and phenotype pages) produced from standardized reanalysis of multi‑omic data across more than 1,000 tumor samples and 10 cancer types.
Zang demonstrated the gene‑centric interface using AKT1 as an example and walked through views that compare tumor versus normal at RNA and protein levels, Manhattan plots for phenotype associations, and tables that expose cohort‑level p‑values and underlying data points for download or downstream enrichment analysis. He described options to retrieve underlying points from meta‑analysis p‑values and to export columns to web tools for gene set analysis.
To illustrate research utility he presented three short case studies. First, an understudied druggable ion‑channel protein (referred to in the talk as COM5) showed strong associations with stromal/EMT and TGF‑β phenotypes and had phosphosites detected across cohorts that were frequently upregulated in tumors, suggesting hypotheses about regulation and kinase linkage (noted correlation with PRKG1).
Second, Zang presented how TP53 mutation status associates differently with protein versus RNA and copy number: some proteins are more strongly altered at the protein level than the RNA level in TP53 mutant tumors, and he highlighted phosphosite regulation on TP53BP1 with kinase associations (CDK1, PRK1) that merit further study.
Third, tumor‑vs‑normal meta‑analyses identified proteins (SMARCA5, UTP4 among examples shown) that are consistently upregulated at the protein but not RNA level, and trans‑association analysis revealed strong protein–protein correlations (for example, SMARCA5 with BAZ1B) and links to molecular phenotype scores (G2/M checkpoint, E2F targets), suggesting protein‑level measures better reflect functional activity for some genes.
Zang emphasized that LinkOmics KB is intended to make harmonized proteogenomics accessible to biologists and clinicians by exposing precomputed molecular phenotype scores and easy searches. He concluded by acknowledging collaborators and the CPTAC pancancer harmonization working group, and opened the session to questions about data export and statistical methods.
Next steps: users interested in the LinkOmics KB resource were directed to the portal (link provided during the seminar) and encouraged to query gene pages and contact the team for access to underlying data tables or clarification of statistical methods.
