The pan-cancer view

Keynote Ilya Shmulevich (Institute for Systems Biology, Seattle) on The Cancer Genome Atlas

IMG_1974For those who want to take a different view towards cancer, browsing The Cancer Genome Atlas (TCGA) may very well generate a host of new ideas and directions that inspire further exploration. During his keynote lecture, Ilya Shmulevich (Institute for Systems Biology, Seattle) proves himself an enthusiastic guide through the still expanding molecular profiling database that is the TCGA. Working with 150 tissue sourcing sites all over the world, patient samples are collected and from each patient two sequences are derived: germ and somatic. The data stored in TCGA go beyond mere sequences, multiple data types are included. As Shmulevich stresses: “No platform left behind.” The overall goal is to analyze 10,000 tumours of more than 30 types of cancer. Which will create a perfect environment for what Shmulevich calls a ‘pan-cancer analysis’. For example, comparing pathways that are associated with a p53 mutation in different types of cancer. More and more we learn that cancers are less determined by the organ they originate in and more by particular genetic and molecular characteristics. Shmulevich: “It appears that basal breast cancer, a number of lung cancers and certain types of ovarian cancer are very similar on the molecular level.” Another example is to take a multilevel pan-cancer view from genes to hallmarks. What is the effect of a mutation in different cancers and which cancers exhibit which hallmarks. “It provides insight into what you could call mutational investment. What is the outcome of the certain mutations, what is the return for the cancer cell?” Needless to say that a huge array of bioinformatics and visualization tools is employed and developed to maximize the use of this incredible collection data. And important to bioinformaticians: the initial raw data remain available to unravel how a particular association has been generated.