AI-Driven Insights in Cancer Genomics

The rapid expansion of genomic technologies has transformed oncology, enabling comprehensive profiling of tumor DNA at unprecedented scale. Yet, the complexity and dimensionality of genomic data, spanning variants, copy number alterations, and structural changes, pose challenges for interpretation and clinical translation. Artificial intelligence (AI) and machine learning are indispensable for uncovering patterns within cancer genomes, predicting disease trajectories, and identifying therapeutic targets. By advancing methods tailored to genomic data, AI can refine molecular classifications, enhance diagnostic accuracy, and guide precision oncology strategies. Recent advances demonstrate the potential of AI-driven genomics to improve early detection, optimize biomarker discovery, and personalize treatment decisions. This special issue highlights cutting-edge research at the intersection of AI and cancer genomics, showcasing novel algorithms, translational applications, and future directions for integrating AI-based genomic insights into routine oncological care.

The goal of this special issue is to assemble innovative research and expert perspectives on AI methods for analyzing cancer genomic data and advancing clinical practice. As tumor genomes generate vast volumes of sequence and variation data, conventional analyses often fall short. AI approaches, from deep learning for variant calling and functional effect prediction to probabilistic models for clonality and evolution, offer powerful solutions to identify meaningful patterns, refine tumor classification, and predict patient outcomes. We will feature work spanning algorithm development for whole-genome/exome data, structural variant discovery, copy-number and mutational signature analysis, and translational pipelines that connect genomic findings to clinical decision-making. Special attention will be given to AI in genomic biomarker discovery, therapy response prediction from genomic features, early detection (including circulating tumor DNA as a genomic analyte), and real-time monitoring using minimally invasive genomic assays. Ultimately, this issue aims to foster dialogue among computational, biological, and clinical communities and to provide a roadmap for integrating AI-driven genomic insights into routine practice to deliver more accurate diagnoses, tailored therapies, and improved outcomes.

This Research Topic welcomes contributions on AI and machine learning for:

o Variant calling, annotation, and prioritization in cancer genomes
o Copy-number, structural variant, mutational signature, and tumor evolution inference
o Genomic risk and outcome prediction; genotype–phenotype modeling
o AI-driven liquid biopsy applied to genomic signals (ctDNA fragmentation, methylation-based genomic signatures framed as genomic context)
o Data quality control, batch correction, and artifact reduction specific to genomic sequencing
o Model interpretability, validation, and clinical deployment for genomics-centric pipelines
o Manuscript types include original research, comprehensive reviews, mini-reviews, perspectives, and methods papers. Interdisciplinary submissions connecting computational sciences, biology, and clinical oncology are especially encouraged.