Fig. 1

Overview of PATOPA. Suppose we are to determine the temporal order of alterations in four pathways, A, B, C and D. The mutation data for pathways A, containing genes A₁, A₂, A₃, and B, containing genes B₁, B₂, B₃, are illustrated in the figure using colors, where a darker color indicates a mutation that is more likely to be functional and a lighter color indicates a mutation that is less likely to be functional based on functional annotation. PATOPA integrates the mutation data, pathway information, and functional annotations to estimate a pivotal probability matrix P, where the (k,i) element of the matrix indicates the probability of the kth functional mutation occurring in the ith pathway. Based on P, we infer the temporal order of the four pathways, where P(A<B) is the probability of A being altered before B and P(B<A) is the probability of B being altered before A. Based on those probabilities, a partial order plot is constructed to show the carcinogenesis process, where the thickness of an edge from A to B is determined by P(A<B)