Research

Interdisciplinary nature of proteogenomics

Proteogenomics: Bridging the Knowledge Gap in Tumor Systems Biology

Cancer is a genetic disease. Genetic abnormalities—such as somatic mutations, copy number alterations, and gene fusions—contribute to aberrant gene regulation and expression, eventually leading to the malignancy-associated dysregulation of cell growth. Over the past decade, large-cohort studies (e.g., TCGA) have mapped the landscape of tumor genetic aberrations, providing a deep understanding of tumor biology. However, our systematic understanding of tumor biology at the protein level has lagged behind. Proteins and PTMs represent not only essential components of tumor biology but also key diagnostic biomarkers and drug targets. Proteogenomics addresses this gap by profiling global protein expression and PTMs using mass spectrometry (MS). By integrating this proteomic data with genomic knowledge, we aim to understand the functional outcomes of genetic abnormalities and guide the targeting of tumor vulnerabilities.

Tumor Pathway Interpretation and Clinical Applications

Most signaling cascades rely heavily on protein-protein interactions and PTMs. Proteogenomics directly measures these components, thereby improving the inference of pathway activity in specific biological contexts.

For example, in our head and neck cancer study, we demonstrated that ligand-dependent EGFR pathway activity could be inferred only through phosphorylation data, not by gene expression (the standard measure in computational biology). We showed that combining these proteomic pathway readouts with genomic abnormalities can identify key biomarkers to guide targeted EGFR mAb treatment.

Proteogenomic profiling of ligand-depend EGFR pathway in head and neck cancer.

We develop and apply computational methods to systematically infer pathway activities using proteogenomic data and explore their clinical applications (e.g., drug response prediction).

Functional Interpretation of Genomic Abnormalities

Genomic studies often identify a long list of tumor genetic aberrations. Integrating these genetic abnormalities with protein and PTM alterations helps discriminate tumor drivers from passengers and prioritize drug targets.

For example, in our colon cancer study, we observed that while copy number amplification occurred across multiple genomic loci, only a subset of these events led to increased protein abundance. Furthermore, these alterations converged on several oncogenic pathways, including the previously under-appreciated endocytosis pathway.

Genome-wide copy number change converges to abnormal endocytosis in colon cancer

We develop and apply computational methods to interpret the functional outcomes of genomic abnormalities using protein and PTM profiling.

Unlocking the Potential of Non-Canonical Peptides and Proteins

While standard proteogenomic analyses rely on known gene annotations, tumor cells harbor genetic abnormalities that generate novel, "non-canonical" proteins undetectable by standard databases. These arise from sources such as somatic mutations, structural variations, hERVs, translation of UTRs, and alternative splicing.

These non-canonical proteins are clinically significant: they drive tumor progression and serve as unique drug targets, particularly as surface-presented neoantigens for immunotherapy.

Detecting these targets requires sophisticated integration of genomics and mass spectrometry. We have previously developed pipelines to identify tumor neoantigens by combining MS proteomics with WES and RNA-seq. Despite this progress, the vast landscape of non-canonical peptides remains largely unexplored, underscoring the critical need for continued computational innovation in the field.

A proteogenomic pipeline to identify tumor neoantigens

We develop and apply computational methods to to identify tumor antigens and proteoforms and understand their biological and clinical roles.

Precision Oncology through Proteogenomics

Tumor progression is driven by molecular mechanisms that vary significantly among patients. We use integrated proteogenomic analysis to stratify patients into molecular subtypes that not only reflect underlying biology but also guide precision therapy. Our research demonstrates that proteogenomic subtyping offers superior clinical relevance compared to genomics or transcriptomics alone. By directly measuring the protein layer, we can identify actionable biomarkers and specific drug targets for distinct patient subgroups.

Patient stratification using proteogenomics-based biomarkers

We perform integrated proteogenomic analyses to understand inter-tumor heterogeneity and connect these insights to precision oncology.

Page updated

Google Sites

Report abuse