Comparative Gene Expression Analysis of Pediatric Leukemias
Mentor: Dr. Olena Vaske
Pediatric cancers typically have lower rates of somatic mutation compared to adult cancers. UCSC Treehouse Childhood Cancer Initiative develops methods for identifying candidate pediatric cancer driver pathways using the analysis of tumor RNA sequencing (RNA-Seq) data. Treehouse currently uses gene expression outlier analysis to identify overexpressed genes in individual tumor samples as compared to a reference compendium of over 12,000 adult and pediatric tumors. This analysis works well for solid tumors but is problematic for leukemias, due to unique expression profiles in the blood compared to other tissues. To address this limitation, a bioinformatic tool called Hydra was developed by a graduate student in Treehouse. Instead of comparing single tumors to the whole data compendium, Hydra identifies disease subtypes and then assigns each sample to a subtype. Hydra analysis of small blue round cell tumors, rhabdomyosarcoma, synovial sarcoma, neuroblastoma, Ewing sarcoma and osteosarcoma revealed several novel molecular subtypes: proliferative signaling, extracellular matrix organization, and immune signaling, shared across the cancer types. We hypothesize that Hydra will be able to identify disease subtypes for leukemia samples and help Treehouse identify cancer driver genes and pathways for these diseases. Hydra outputs data on several thousand pathways and several hundred genes. These data must be thoroughly investigated, curated and critically analyzed by a researcher. This work on the Hydra project will improve Treehouse’s ability to analyze pediatric leukemias and could also discover novel disease subtypes which may help find therapies for patients who relapse or do not respond to initial treatment.
