Decoding Tumour Microenvironments in Xenograft Models Using Unsupervised Learning of Spatial Transcriptomics Data
Identification
- Clé publique
- radar:r-f9d6951e5be473a0bc89dda5f819c54a
- Auteurs
- Abdelkareem, Aly O.
- Année déclarée
- 2025
- Langue déclarée
- en
- Type source
- Other
- Première détection
- 2026-11-13T08:57:06.011867+00:00
- Dernière observation des métadonnées
- 2026-11-13T08:57:06.011867+00:00
- Dernière vérification HTTP
- Non effectuée
Résumé documentaire — langue originale
x Cancer remains one of the leading causes of mortality worldwide, with glioblastoma (GBM) standing out as an especially aggressive and treatment-resistant malignancy. A central driver of this resistance lies in the tumor microenvironment (TME)—a complex, spatially organized ecosystem of diverse cell types and signaling pathways. Understanding cellular heterogeneity and intercellular communication within the TME is crucial for developing more effective therapies. Patient-derived xenograft (PDX) models, in which human tumors are engrafted into immunocompromised mice, preserve essential tumor characteristics and provide a valuable platform for studying TME dynamics. However, their cross-species nature introduces analytical challenges, especially in resolving cell-type mixtures and interpreting ligand-receptor (LR) signaling across species. This thesis presents two major computational innovations to overcome these challenges using spatial transcriptomics (ST) data. First, it introduces spOT-NMF, a novel unsupervised deconvolution method that integrates Optimal Transport with Non-negative Matrix Factorization. In contrast to popular reference-based approaches that depend on single-cell reference datasets—often unavailable or incomplete in complex tissues—spOT-NMF is fully reference-free. It disentangles mixed gene expression profiles among samples by leveraging the Wasserstein distance, enabling accurate and scalable analysis across varying spatial resolutions and sequencing depths. Extensive benchmarking against established unsupervised methods such as LDA, NMF, and cNMF demonstrates that spOT-NMF consistently outperforms these models, particularly in low-signal and heterogeneous xenograft settings. Second, a computational framework (XenoSignal) is developed to discover and enhance interpretability of inter- and intraspecies LR interactions in xenografts. Since existing LR inference tools are constrained to single-species databases, typical analytic strategies for xenograft data rely on human-mouse orthologues and on the strong assumption that such orthologues have minimal functional divergence. Here, an analytic pipeline was established to determine the subset of proteins for which this assumption holds or fails. Leveraging AlphaFold3 to predict three-dimensional protein structures, XenoSignal quantifies interaction plausibility between human ligands and murine receptors, and vice versa. This strategy enables the construction of a structurally annotated cross-species LR interaction atlas, facilitating the discovery of intercellular communication pathways relevant to tumor progression and immune evasion. Together, these contributions advance the computational toolkit for spatial transcriptomics, with a particular focus on xenograft models. By enhancing deconvolution accuracy and enabling interspecies signaling inference, this work supports a deeper understanding of the spatial and molecular architecture of the TME. These methods lay the groundwork for future efforts in precision oncology, including the identification of therapeutic targets and the design of context-aware interventions in complex tumor ecosystems.
Identifiants
- DOI
- 10.11575/prism/50507
- CITATION
- Abdelkareem, A. O. (2025). Decoding tumour microenvironments in xenograft models using unsupervised learning of spatial transcriptomics data (Doctoral thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca.
- URI
- https://hdl.handle.net/1880/122913
Repères documentaires
- Famille documentaire
- unknown
- Types déclarés par les sources
- Other
- Éditeurs
- Cumming School of Medicine
Les mentions d’entités ne constituent pas une preuve de reconnaissance juridique. La juridiction étudiée est distincte du pays d’affiliation des auteurs.
Sources
DataCite · Document 1
Citer cette référence
Abdelkareem, Aly O.. Decoding Tumour Microenvironments in Xenograft Models Using Unsupervised Learning of Spatial Transcriptomics Data. 2025. https://doi.org/10.11575/prism/50507
JSON · CSL-JSON · BibTeX · RIS · CSV · JSON-LD ·
APA, Chicago, ISO 690 : formats bibliographiques fondés sur les champs disponibles. OSCOLA / Bluebook : données structurées uniquement ; aucune conformité juridique complète n’est revendiquée sans vérification des éléments nécessaires.
- APA
- Abdelkareem, Aly O.. (2025). Decoding Tumour Microenvironments in Xenograft Models Using Unsupervised Learning of Spatial Transcriptomics Data. https://doi.org/10.11575/prism/50507
- Chicago
- Abdelkareem, Aly O.. Decoding Tumour Microenvironments in Xenograft Models Using Unsupervised Learning of Spatial Transcriptomics Data. 2025. https://doi.org/10.11575/prism/50507
- ISO 690
- Abdelkareem, Aly O.. Decoding Tumour Microenvironments in Xenograft Models Using Unsupervised Learning of Spatial Transcriptomics Data. 2025. https://doi.org/10.11575/prism/50507
Warning: Undefined array key "last_link_verified" in /home/clients/e27d81e12f04d2bfd747aa264540ac56/sites/_UDTR_APP/legal_library_reference.php on line 31
UDTR — Bibliothèque juridique mondiale des droits de la Nature. Decoding Tumour Microenvironments in Xenograft Models Using Unsupervised Learning of Spatial Transcriptomics Data. Notice documentaire: https://www.declarationuniverselledesdroitsdelarbre.org/global-rights-of-nature-radar/reference/r-f9d6951e5be473a0bc89dda5f819c54a/ · Dernière vérification HTTP : non effectuée
Données de cette version figée
Dossiers et relations documentaires
Rôles des sources
Source bibliographique / publication d’origine · https://ucalgary.scholaris.ca/handle/1880/122913
Un lien vers une autorité ne prouve pas à lui seul une affirmation juridique. Les résumés et les textes intégraux restent soumis aux conditions de leur source.
Statut documentaire
Pertinence à examiner
Bibliothèque
Cette identité n’est pas encore intégrée à la Bibliothèque juridique.