Peptide-based Clustering for Determination of Colorectal Cancer Subtypes

Bioinformatics Internship Presentation

Amina Jackson (Mentor: Dr. Nathan Edwards, Department of Biochemistry and Molecular & Celluar Biology, Georgetown University)

September 1st, 2015, 4:00-4:20pm, Room 1300, Harris Building

Characterization of genomic features is fundamental to the understanding of the molecular mechanism of our biological systems and the mechanics of different diseases. We believe peptide abundance can be an added alternative and could improve some of bioinformatics errors associated with the currently used protein abundance and mRNA transcript abundance to identify these genomic features. However, we focused our project on the use of peptide-based quantification to infer the different cancer subtypes by clustering. We followed the hypothesis: can peptide based quantification by clustering be used to infer cancer subtypes and produce similar subtype assignments as published protein based and mRNA based cancer subtypes. We used Cancer Genome Atlas (TCGA) data and GenePattern consensus clustering tool used for colorectal cancer characterization (CRC) in Zhang et al (2014). We were successful in generating similar clusters of samples using peptide based clustering, as did Zhang et al using protein based clustering.

Zhang et al (2014). Proteogenomic characterization of human colon and rectal cancer . Nature. , 513 (7518), 382-387.