Soft-Gen: Software e dispositivi NGS per la determinazione genetica in clinica

Next-generation sequencing (NGS) has transformed genomics by enabling high-throughput characterization of genetic variation, with major impact on research and clinical practice. Among structural variants, copy number alterations (CNAs), unbalanced DNA segmental changes ≥50 kb, are particularly relevant as they contribute to neurodevelopmental, autoimmune, and psychiatric disorders, and are a hallmark of cancer. In tumors, somatic CNAs (SCNAs) drive oncogene activation, tumor suppressor loss, and genomic instability, influencing disease progression and therapy response. Despite their importance, CNA detection in clinical settings remains challenging. Whole-genome sequencing (WGS) provides comprehensive profiling but is limited by cost, data burden, and computational requirements. In contrast, shallow whole-genome sequencing (sWGS), based on ultra-low sequencing depth (<1×), enables accurate and cost-effective CNA detection while minimizing storage and processing needs, making it particularly suitable for translational applications such as liquid biopsy. The central focus of my doctoral research, conducted within the SmartPhD program (University of Padova and AB Analitica S.r.l.), was the development of robust bioinformatics methodologies to harness sWGS data for CNA profiling in clinical contexts. The major outcome was SAMURAI (Shallow Analysis of Copy Number Alterations Using a Reproducible and Integrated pipeline), the first publicly available, modular workflow specifically designed for sWGS CNA analysis. SAMURAI integrates state-of-the-art algorithms for CNA detection and post-processing, supports both solid and liquid biopsy samples, and ensures reproducibility across datasets and applications. By addressing the lack of accessible and validated workflows, it provides a crucial resource for research and clinical translation. The translational potential of SAMURAI was demonstrated by applying the pipeline to two independent patient cohorts: (i) plasma-derived cell-free DNA from small cell lung cancer (SCLC) patients, and (ii) hepatocellular carcinoma (HCC) samples. In both cases, SAMURAI enabled detection of clinically relevant CNA patterns, supporting liquid biopsy as a minimally invasive tool for tumor diagnosis, disease monitoring, and prognostic assessment. These studies underscored the capacity of sWGS to capture tumor dynamics in real time and highlighted its value as a complement to tissue biopsies. We extended SAMURAI to explore additional biomarkers. In particular, we investigated the detection of homologous recombination deficiency (HRD) from sWGS data. HRD, the inability to repair DNA double-strand breaks via homologous recombination, is an actionable biomarker predictive of response to PARP inhibitors and related therapies. By applying methods to extract HRD-associated genomic signatures from low-coverage data, SAMURAI was expanded beyond CNA profiling into therapeutic vulnerability prediction. To promote adoption in translational and clinical settings, where bioinformatics expertise may be limited, a graphical user interface (GUI) was developed as the final step of the project, making SAMURAI user-friendly and lowering technical barriers to integration into diagnostic workflows. In summary, this doctoral research demonstrates how sWGS, coupled with standardized bioinformatics solutions, can bridge the gap between genomic research and clinical application. By enabling accurate CNA and HRD detection, SAMURAI provides a scalable framework for advancing personalized medicine, supporting patient stratification, disease monitoring, and therapeutic decision-making. Beyond oncology, its modular design also holds promise for diverse areas of clinical genomics, including non-invasive prenatal testing. Collectively, this work emphasizes the pivotal role of bioinformatics innovation in translating sequencing technologies into actionable clinical tools.