Untangling a complex web: Computational analyses of tumor molecular profiles to decode driver mechanisms

Abstract

Genome-scale studies focusing on molecular profiling of cancers across tissue types have revealed a plethora of aberrations across the genomic, transcriptomic, and epigenomic scales. The significant molecular heterogeneity across individual tumors even within the same tissue context complicates decoding the key etiologic mechanisms of this disease. Furthermore, it is increasingly likely that biologic mechanisms underlying the pathobiology of cancer involve multiple molecular entities interacting across functional scales. This has motivated the development of computational approaches that integrate molecular measurements with prior biological knowledge in increasingly intricate ways to enable the discovery of driver genomic aberrations across cancers. Here, we review diverse methodological approaches that have powered significant advances in our understanding of the genomic underpinnings of cancer at the cohort and at the individual tumor scales. We outline the key advances and challenges in the computational discovery of cancer mechanisms while motivating the development of systems biology approaches to comprehensively decode the biologic drivers of this complex disease.

Keywords: Functional impact; Multiomics integration; Mutational significance; Mutations; Pan-cancer analysis; Systems biology.

Fig. 1.

Computational approaches to prioritize cancer driver genes and pathways categorized into major methodological categories: Analysis of Mutation Rate; Analysis of Protein Sequence, Structure and Function; Integrative Analysis of Multi-Omics Measurements; and Analysis of Whole Genomes. Also shown are the key computational tools within each category.

Fig. 2.

Schematic representation of the mutation rate-based approaches. BMR is estimated per gene (G_i) across all sequenced bases in the cohort of samples, and scores (N_i) represent the number of samples harboring mutations in G_i. In Gene Level mutation rate methods compare N_i with a significance threshold derived from the background mutation rate (BMR). If N_i ≫ BMR, G_i is identified as a significantly mutated gene (SMG); In contrast, Subnetwork Level methods evaluate BMR and N_i within a PPI network, where red and blue colors in the piecharts within each gene’s node indicate the proportion of mutated vs. wild-type samples. These methods then propagate the mutational frequencies throughout the network, using insights from heat diffusion processes, thus determining connected sub-networks that are more frequently targeted for mutations than expected.

Fig. 3.

Shown are the three major categories of methods that utilize either protein sequence homology, protein structure, and functional hotspots. A: Sequence homology-based methods are based on the insight that mutations in conserved regions (green box) are more likely to be deleterious as compared to the mutations in the non-conserved regions (red box). B: Protein structure- based methods not only utilize sequence homology but also model changes in the structure of the wild-type (WT) and mutated (MUT) protein. C: Since functional hotspots play an important role in protein-protein/protein-ligand interactions, this figure shows a close up view of the ligand binding pocket and how a mutation impacts the interaction of the protein with the ligand. The figure shows overlapping structures of the wildtype (blue) and mutated (red) proteins in addition to depicting the disrupted ligand interaction.

Fig. 4.

Schematic representation of the Multi-Omics integrative strategies. These methods integrate Multi-omics Data such as expression changes (RNA sequencing), copy number alterations (CNA), Mutation information (SNV) and epigenetic alterations (DNA Methylation) with Prior Biological Information such as PPI networks to estimate functional impacts of mutated genes on their downstream target genes. The Functional Impact of Mutated Genes, for example gene G_i, is evaluated by assessing for differential expression of downstream target genes {G_t1,G_t2….,G_tn} in mutated vs wild-type samples of G_i. These integrative analyses result in subnetworks containing mutated genes with significant impact on downstream targets. Note that the red and blue colors in the pie-charts within each mutated gene’s node indicate the proportion of mutated vs. wild-type samples; the shade of red indicates impact level of mutated gene G_i to the G_ti.