Advances in next-generation sequencing have provided high-dimensional RNA-seq datasets, allowing the stratification of some tumor patients based on their transcriptomic profiles. Machine learning methods have been used to reduce and cluster high-dimensional data. Recently, uniform manifold approximation and projection (UMAP) was applied to project genomic datasets in low-dimensional Euclidean latent space. Here, we evaluated how different representations of the UMAP embedding can impact the analysis of breast cancer (BC) stratification. We projected BC RNA-seq data on Euclidean, spherical, and hyperbolic spaces, and stratified BC patients via clustering algorithms. We also proposed a pipeline to yield more reproducible clustering outputs. The results show how the selection of the latent space can affect downstream stratification results and suggest that the exploration of different geometrical representations is recommended to explore data structure and samples’ relationships.

Next Generation Sequencing (NGS) has revolutionized genome biology, enabling the rapid sequencing of an entire human genome and facilitating the integration of Whole Genome Sequencing (WGS) into both research and clinical applications. The high-throughput nature of NGS and the complex data processing required has driven the need for advanced computational infrastructures to analyse these large datasets. The aim of this work is to introduce an innovative bioinformatic pipeline, named GeNePi, for the efficient and precise analysis of WGS short paired-end reads. Built on the Nextflow framework with a modular structure, GeNePi incorporates GPU-accelerated algorithms and supports multiple workflow configurations. The pipeline automates the extraction of biologically relevant insights from raw WGS data, including: disease-related variants such as single nucleotide variants (SNVs), small insertions or deletions (INDELs), copy number variants (CNVs), and structural variants (SVs). Optimized for high-performance computing (HPC) environments, it takes advantage of job-scheduler submissions, parallelised processing, and tailored resource allocation for each analysis step. Tested on synthetic and real datasets, GeNePi accurately identifies genomic variants, with performances comparable to that of state-of-art tools. These features make GeNePi a valuable instrument for large-scale analyses in both research and clinical contexts, representing a key step towards the establishment of National Centers for Computational and Technological Medicine.