Abstract:Tumor genome sequencing leads to documenting thousands of DNA mutations and other genomic alterations. At present, these data cannot be analyzed adequately to aid in the understanding of tumorigenesis and its evolution. Moreover, we have little insight into how to use these data to predict clinical phenotypes and tumor progression to better design patient treatment. To meet these challenges, we discuss a cancer hallmark network framework for modeling genome sequencing data to predict cancer clonal evolution and associated clinical phenotypes. The framework includes: (1) cancer hallmarks that can be represented by a few molecular/signaling networks. 'Network operational signatures' which represent gene regulatory logics/strengths enable to quantify state transitions and measures of hallmark traits. Thus, sets of genomic alterations which are associated with network operational signatures could be linked to the state/measure of hallmark traits. The network operational signature transforms genotypic data (i.e., genomic alterations) to regulatory phenotypic profiles (i.e., regulatory logics/strengths), to cellular phenotypic profiles (i.e., hallmark traits) which lead to clinical phenotypic profiles (i.e., a collection of hallmark traits). Furthermore, the framework considers regulatory logics of the hallmark networks under tumor evolutionary dynamics and therefore also includes: (2) a self-promoting positive feedback loop that is dominated by a genomic instability network and a cell survival/proliferation network is the main driver of tumor clonal evolution. Surrounding tumor stroma and its host immune systems shape the evolutionary paths; (3) cell motility initiating metastasis is a byproduct of the above self-promoting loop activity during tumorigenesis; (4) an emerging hallmark network which triggers genome duplication dominates a feed-forward loop which in turn could act as a rate-limiting step for tumor formation; (5) mutations and other genomic alterations have specific patterns and tissue-specificity, which are driven by aging and other cancer-inducing agents. This framework represents the logics of complex cancer biology as a myriad of phenotypic complexities governed by a limited set of underlying organizing principles. It therefore adds to our understanding of tumor evolution and tumorigenesis, and moreover, potential usefulness of predicting tumors' evolutionary paths and clinical phenotypes. Strategies of using this framework in conjunction with genome sequencing data in an attempt to predict personalized drug targets, drug resistance, and metastasis for cancer patients, as well as cancer risks for healthy individuals are discussed. Accurate prediction of cancer clonal evolution and clinical phenotypes will have substantial impact on timely diagnosis, personalized treatment and personalized prevention of cancer.

From genome to phenome: Predicting multiple cancer phenotypes based on somatic genomic alterations via the genomic impact transformer

Identification of Tumor Mutation Burden, Microsatellite Instability, and Somatic Copy Number Alteration Derived Nine Gene Signatures to Predict Clinical Outcomes in STAD.

Transformer for Gene Expression Modeling (T-GEM): An Interpretable Deep Learning Model for Gene Expression-Based Phenotype Predictions

MGCT: Mutual-Guided Cross-Modality Transformer for Survival Outcome Prediction using Integrative Histopathology-Genomic Features

Bioinformatics Analysis of Thousands of TCGA Tumors to Determine the Involvement of Epigenetic Regulators in Human Cancer

Cancer-inspired Genomics Mapper Model for the Generation of Synthetic DNA Sequences with Desired Genomics Signatures

Predicting Gene Spatial Expression and Cancer Prognosis: An Integrated Graph and Image Deep Learning Approach Based on HE Slides

Prediction of Cancer Patient Responses to Different Treatments Using Personalized Tumor Models or Models with Matching Genomic Profiles.

Predictive genomics: A cancer hallmark network framework for predicting tumor clinical phenotypes using genome sequencing data

Deep Learning Model for Tumor Type Prediction using Targeted Clinical Genomic Sequencing Data

Biological Informed Graph Neural Network for Tumor Mutation Burden Prediction and Immunotherapy-Related Pathway Analysis in Gastric Cancer

Integrating gene mutation spectra from tumors and the general population with gene expression topological networks to identify novel cancer driver genes

A Multi-Omics Integration Framework Using Graph Attention Networks for Cancer Subtype Prediction

Multifactorial Deep Learning Reveals Pan-Cancer Genomic Tumor Clusters with Distinct Immunogenomic Landscape and Response to Immunotherapy.

Identification of Cancer Driver Genes by Integrating Multiomics Data with Graph Neural Networks

HGTDG: An Interpretable Heterogeneous Graph Transformer Framework for Cancer Driver Gene Prediction.

Advancing cancer driver gene detection via Schur complement graph augmentation and independent subspace feature extraction

A heterogeneous graph transformer framework for accurate cancer driver gene prediction and downstream analysis

MODIG: Integrating Multi-Omics and Multi-Dimensional Gene Network for Cancer Driver Gene Identification Based on Graph Attention Network Model.

Harnessing TME depicted by histological images to improve cancer prognosis through a deep learning system

MetaGXplore: Integrating Multi-Omics Data with Graph Convolutional Networks for Pan-cancer Patient Metastasis Identification