Xiaojing Wang, Ph.D.
Our research focuses on developing novel computational tools for integrative analyses of proteomic and genomic data to help advance our understanding of cancer biology.
- Computational Biology
- NGS data analysis
- Genomic data analysis
- Data visualization
- A major interest for us is proteogenomics, an emerging field that aims to characterize proteins in the context of genomic changes, which are common in complex diseases such as cancer. Building upon our previous studies, we are working on predicting cancer neoantigens and the adaptive immune repertoires using proteomic data. Accumulating evidence shows the significance of anti-tumor immune responses in predicting the prognosis of cancer patients. Currently, somatic mutations are commonly used to predict neoantigens. However, due to post translational modifications and selective transcription of mutant or wild type alleles, mutations do not provide the most immediate information for neoantigen prediction. Instead, the definitive proof of a mutation triggering an immune response is the detection of the encoded protein/peptide. Our project is poised to investigate the correlation between proteomic profiles and response to immunotherapy, a therapeutic modality that shows promises in a number of cancer types including lung cancer, melanoma, and MSI colorectal cancer. Our goal is to explore the feasibility of using proteomics data to integrate the latest advance of cancer genomics into proteomics and prioritize neoantigen and further promote proteogenomic integration in this field.
- Building oncogenic map of childhood cancers. Matching the genetic profile of a tumor to the right medicine is the cornerstone of precision oncology. This is in part achieved in adults through techniques such as panel sequencing. However, similar approach is not suitable for children with cancer because unlike adults, childhood cancers typically have much less mutations thus much less targetable genetic changes. We are building an oncogenic map by training computational models from adult data and applying them to childhood cancers. The novelty of this approach is that the resulting model is from real patient data thus fully capture the heterogeneity of the tumors. In addition, once the model is built, it no longer relies on genetic data thus application to children will not be limited by their sparse genetic alterations.