Salvi, Marcello (2024) Translating Transcriptomic Insights into Precision Medicine: A Roadmap for CUP and Breast Cancer Stratification. [Tesi di dottorato]
![[thumbnail of SalviMarcello_thesis_CQB_XXXVII.pdf]](https://www.fedoa.unina.it/style/images/fileicons/other.png "SalviMarcello_thesis_CQB_XXXVII.pdf") |
Documento PDF
SalviMarcello_thesis_CQB_XXXVII.pdf Visibile a [TBR] Utenti registrati Download (8MB) | Richiedi una copia |
|
Documento PDF
Salvi_Marcello_XXXVII_PARZIALE.pdf Visibile a [TBR] Amministratori dell'archivio Download (5MB) | Richiedi una copia |
| Tipologia del documento: | Tesi di dottorato |
|---|---|
| Lingua: | English |
| Titolo: | Translating Transcriptomic Insights into Precision Medicine: A Roadmap for CUP and Breast Cancer Stratification |
| Autori: | Autore Email Salvi, Marcello marcellosalvi96@gmail.com |
| Data: | 12 Dicembre 2024 |
| Numero di pagine: | 104 |
| Istituzione: | Università degli Studi di Napoli Federico II |
| Dipartimento: | Ingegneria Elettrica e delle Tecnologie dell'Informazione |
| Dottorato: | Computational and quantitative biology |
| Ciclo di dottorato: | 37 |
| Coordinatore del Corso di dottorato: | nome email CECCARELLI, MICHELE michele.ceccarelli@unina.it |
| Tutor: | nome email Cacchiarelli, Davide [non definito] |
| Data: | 12 Dicembre 2024 |
| Numero di pagine: | 104 |
| Parole chiave: | : Transcriptomics, 3' Digital RNA-seq, Machine Learning, Breast Cancer, CUP |
| Settori scientifico-disciplinari del MIUR: | Area 05 - Scienze biologiche > BIO/11 - Biologia molecolare |
| Informazioni aggiuntive: | Necessito secretazione tesi 36 mesi poichè i risultati sono sotto brevetto da parte di NEGEDIA srl |
| Depositato il: | 18 Nov 2025 11:54 |
| Ultima modifica: | 09 Ago 2026 05:59 |
| URI: | https://www.fedoa.unina.it/id/eprint/16428 |
Abstract
Background: Traditional oncology diagnostics, such as immunohistochemistry (IHC), often exhibit limitations in sensitivity, reproducibility, and quantitative outputs. Transcriptomics and next-generation sequencing (NGS) technologies can enhance diagnostic accuracy by offering detailed, quantitative gene expression data. By extracting value from these data with a Machine learning algorithm, we can better dissect particularly complex heterogeneous cases like breast cancer and carcinoma of unknown primary (CUP). Objective: To develop and validate a streamlined bioinformatics workflow using 3' RNA-seq for robust, quantitative gene expression analysis in formalin-fixed paraffin-embedded (FFPE) tissue samples. The goal is to enhance diagnostic precision in breast cancer, quantify key biomarkers per ESMO recommendations, and apply ML-driven tissue origin prediction for CUP cases. Methods: We implemented 3' RNA-seq, allowing reproducible measurements of over 10,000 genes from degraded FFPE RNA. The quantification of biomarkers (ER, PR, HER2, Ki-67) aligned with ESMO guidelines. Advanced ML algorithms were employed to predict tissue origin in CUP, utilising transcriptomic data from various sample types, including patient-derived xenografts (PDx) and agnospheres. Model performance was compared with IHC across all samples. Results: The workflow generated high-quality gene expression profiles from FFPE samples. In breast cancer, molecular quantification provided consistent biomarker profiles, improving therapeutic decision-making. In CUP, ML techniques outperformed IHC, classifying more samples into tissue-specific categories. Discrepancies between ML and IHC highlighted limitations in 6 traditional diagnostics, with the models confidently classifying ambiguous samples, thus improving accuracy.
Downloads
Downloads per month over past year
Actions (login required)
 |
Modifica documento |
