Ben Brett will Present His Research on August 22, 2013
Purpose: Cancer is a complex disease caused by many somatic mutations. Cancer is often the result of random mutations. Random mutations can be classified as driver (causative) and passenger (non-causative) mutations. Differentiation between the driver mutations and passenger mutations is extremely difficult in human cancers due to the complexity of the disease and mutations. In addition, understanding the mechanisms of these genes is confounded by the rarity of many cancer genes. Network biology has been used to attempt to understand the relationship between genes and the possible shared functions of these genes. However, these networks are incomplete and often only contain well-characterized connections, making it difficult to identify more distantly connected genes.
Methods: A relatedness network was created utilizing network reliability algorithms. The relatedness of genes is the reliability of the connection between two genes. This was determined using the Edge-Packing Bounds algorithm with maximum of 4 intermediate genes, maximum of 5 paths, and minimum edge weight of 50%. This network was used to determine the relatedness between genes. By combining multiple independent paths we can better visualize and identify the connections between two genes.
Discussion: By combining these independent paths, we increase the number of pairwise comparisons from ~518,000 to ~51 million. This is mainly due to the addition of indirect connections between genes. When we use this information on pathways, we see an increase of strongly connected genes.