The main objective of the current Ph.D. dissertation was the implementation of a computational workflow, which integrates high- throughput methodologies already available, to improve the piRNA analysis, with the aim to decipher the putative association of piRNAs with complex diseases/developmental processes. Indeed, the vast majority of current tools and pipelines, suffer from major methodological pitfalls regarding the quantification and annotation of piRNAs, such as the use of outdated piRNA databases, inconsistent piRNA IDs, discrepancies between sncRNA classes in the annotation of the same sequences, etc. On this premise, a start-to-end analytical workflow was developed, in order to mitigate piRNA annotation inconsistencies and provide the scientific community with a methodological framework for the analysis of smallRNA-seq data with a focus on piRNAs. Through this pipeline, various datasets related to colorectal cancer, mouse cardiomyocytes and cardiac stem cells were investigated to identify piRNAs related to the biological processes in disease progression and stem cell differentiation. The main rationale of the realised workflow was that it can be robustly utilized for the discovery of putative piRNAs that could serve as potential biomarkers (with clinical prognostic and diagnostic utility) and study their association with known oncogenic pathways and stemness processes. Finally, as most of the methods used in this dissertation are algorithmic pipelines that follow the Open Science frameworks, can be found publicly available in the GitHub repository ( https://github.com/ConYel/Dissertation/tree/v1.0.0 ) in which all workflows for all datasets have been published.