Student: Robert Schultz
Project Mentors: Dr. Christopher Plaisier - SBHSE
Dr. Vincent Pizziconi - SBHSE
Dr. Kenneth Buetow - SOLS
YouTube Link: View the video link below before joining the zoom meeting
Zoom Link: https://asu.zoom.us/j/95082512969
Abstract
Advances in genomics technologies have enabled studies of the somatic mutation landscape of tumors. The two most prevalent mutation types are Protein Affecting Mutations (PAMs) where one or a few nucleotides are mutated in a protein coding region of a single gene, and Copy Number Alterations (CNAs) where large segments of DNA are amplified or deleted affecting one or more genes. There are methods that integrate both PAMs and CNAs to discover oncogenic somatic mutations for tumor types. However, no methods currently exist to integrate PAM and CNA somatic mutations into a single more complete mutational profile that can be used to increase power to detect genetic associations in downstream studies. Here, we present OncoMerge, a novel method to systematically combine PAMs with CNAs into a more complete mutational profile for downstream applications by labeling genes as Activating or Loss of Function mutations and filtering those that do not pass low-amplitude coincidence thresholds, mutational frequency thresholds, and a permutation analysis. OncoMerge is used to build the mutational landscape of 32 tumor types using data from TCGA. Among these tumor types 2,106 mutational profiles are expanded, 1,187 are categorized as Activating mutations, and 919 are categorized as Loss of Function mutations. OncoMerge also reveals 556 mutations which are only frequently mutated when protein affecting mutations and copy number alterations are considered together but undercut when separate. The OncoMerged mutational landscapes for each tumor type were examined with hypergeometric under enrichment analysis to reveal at least 13 tumor types (Bonferroni corrected p-value ≤ 0.00161) cannot be significantly represented using one data type alone and only three tumor types can. In conclusion, evaluating only PAMs or CNAs alone will deliver an underrepresented mutational landscape for many tumor types but OncoMerge empowers researchers to overcome this important obstacle.