Volume 10 Number 1 2026
Protein Language Models for Predicting BRCA1/BRCA2 Variant Pathogenicity: A Computational Bridge Toward Precision Oncology
Ghaith Kamil Jawad 1*, Abdulsamie Hassan Alta'ee 2, Oday Jasim Alsalihi 3
Integrative Biomedical Research 10 (1) 1-8 https://doi.org/10.25163/biomedical.10110923
Submitted: 17 May 2026 Revised: 08 July 2026 Accepted: 17 July 2026 Published: 19 July 2026
Breast and ovarian cancers linked to pathogenic BRCA1 and BRCA2 alterations impose a substantial and growing global health burden, and the clinical utility of PARP-inhibitor therapy now depends almost entirely on correctly telling a harmful variant apart from a harmless one. That task has become harder, not easier, as sequencing volumes have grown: nearly four in ten variants identified through clinical gene panels are still classified as variants of uncertain significance (VUS), leaving clinicians and patients in a difficult holding pattern. We conducted a structured narrative synthesis of peer-reviewed literature addressing protein language models (pLMs), multi-omics deep learning architectures, and their translational application to BRCA variant interpretation and precision oncology, following a reproducible search-and-screening workflow across major biomedical databases. Evidence was extracted, tabulated, and thematically organized across four domains: representation learning and data fusion, clinical risk stratification, synthetic-lethality target discovery, and translational barriers to clinical deployment. Across the reviewed literature, pLM- and transformer-based architectures (including AlphaMissense, ESM3, DNABERT-S, SetQuence, and SetOmic) consistently outperformed classical machine learning baselines in variant- and tumor-classification tasks, with several multi-omics fusion frameworks (e.g., SetOmic, MOGONET, TMO-Net) achieving accuracy or F1-scores exceeding 0.90 in pan-cancer and breast-cancer cohorts. Graph-based synthetic-lethality models (DGIB4SL, KR4SL, MAGICAL) further extended this predictive power to therapeutic target discovery beyond canonical BRCA-PARP biology. However, persistent obstacles — dataset homogeneity, limited model interpretability, and underrepresentation of non-European ancestries — continue to restrict real-world generalizability. Protein language models represent a genuinely transformative, though not yet fully mature, tool for resolving BRCA variant ambiguity and expanding equitable access to precision oncology; their clinical adoption will likely hinge on interpretability, prospective validation, and deliberate correction of demographic bias in training data.
Keywords: BRCA1/BRCA2; protein language models; variant pathogenicity prediction; precision oncology; homologous recombination deficiency; PARP inhibitors; multi-omics data integration
Abramson, J., Adler, J., Dunger, J., Evans, R., Green, T., Pritzel, A., ... & Hassabis, D. (2024). Accurate structure prediction of biomolecular interactions with AlphaFold 3. Nature, 630(8015), 211–218. https://doi.org/10.1038/s41586-024-07487-w
Barua, S., Badhrinarayanan, B., & Balaji, S. (2026). Bridging diagnosis and therapeutics: The role of AI in cancer detection and drug development. Health Sciences Review, 19, 100270. https://doi.org/10.1016/j.hsr.2026.100270
Bilal, A., Imran, A., & Baig, T. I. (2024). Breast cancer diagnosis using support vector machine optimized by improved quantum inspired grey wolf optimization. Scientific Reports, 14, 1–25. https://doi.org/10.1038/s41598-024-61322-w
Corso, G., et al. (2026). Leave no data behind: Exploring a new paradigm in oncology with foundation models and large language models. Cell Reports Medicine, 7, 102966. https://doi.org/10.1016/j.xcrm.2026.102966
Dehkharghanian, T., Bidgoli, A. A., Riasatian, A., Mazaheri, P., Campbell, C. J. V., Pantanowitz, L., ... & Tizhoosh, H. R. (2023). Biased data, biased AI: Deep networks predict the acquisition site of TCGA images. Diagnostic Pathology, 18, 67. https://doi.org/10.1186/s13000-023-01355-3
Eniu, A., Pop, L., Stoian, A., Dronca, E., Matei, R., Ligtenberg, M., Ouchene, H., Onisim, A., Rotaru, O., Eniu, R., et al. (2023). Delivering precision medicine in hereditary breast cancer: NGS-based multi-gene panel testing beyond BRCA1/2. Biomedicines, 11(5), 1386. https://doi.org/10.3390/biomedicines11051386
Gentile, G., et al. (2026). Integration of germline testing into precision oncology frameworks: A systematic review. Critical Reviews in Oncology/Hematology, 223, 105307. https://doi.org/10.1016/j.critrevonc.2026.105307
Ismail, T., Alzneika, S., Riguene, E., Al-Maraghi, S., Alabdulrazzak, A., Al-Khal, N., & Nomikos, M. (2024). BRCA1 and its vulnerable C-terminal BRCT domain: Structure, function, genetic mutations and links to diagnosis and treatment of breast and ovarian cancer. Pharmaceuticals, 17(3), 333. https://doi.org/10.3390/ph17030333
Jurenaite, N., León-Periñán, D., Donath, V., Torge, S., & Jäkel, R. (2024). SetQuence & SetOmic: Deep set transformers for whole genome and exome tumour analysis. BioSystems, 235, 105095. https://doi.org/10.1016/j.biosystems.2023.105095
Kotsifaki, A., Kalouda, G., Karalexis, E., Stathaki, M., Metaxas, G., & Armakolas, A. (2025). Emerging breast cancer subpopulations: Functional heterogeneity beyond the classical subtypes. International Journal of Molecular Sciences, 26, 11599. https://doi.org/10.3390/ijms262311599
Koyun, O. C., et al. (2026). AnchorMIL: Multiple instance learning framework with anchored regression for Oncotype DX recurrence score prediction. Expert Systems with Applications, 303, 130469. https://doi.org/10.1016/j.eswa.2025.130469
Kuenzi, B. M., Park, J., Fong, S. H., Sanchez, K. S., Lee, J., Kreisberg, J. F., ... & Ideker, T. (2020). Predicting drug response and synergy using a deep learning model of human cancer cells. Cancer Cell, 38(5), 672–684. https://doi.org/10.1016/j.ccell.2020.09.014
Lash, S., & Valero, C. (2026). Identification of a 17-gene predictive signature through ensemble machine learning analysis for predicting neoadjuvant chemotherapy response in solid tumors. Current Issues in Molecular Biology, 48(1), 94–113. https://doi.org/10.3390/cimb48010094
Li, J., Li, Y., & Xie, T. (2026). Bridging realms: Artificial intelligence integrates omics, generative models, and traditional medicine for anticancer drug innovation. Journal of Pharmaceutical Analysis, 16(1), 85–96. https://doi.org/10.1016/j.jpha.2026.101630
Liang, H., Luo, H., Sang, Z., Jia, M., Jiang, X., Wang, Z., ... & Zhang, Z. (2024). GREMI: An explainable multi-omics integration framework for enhanced disease prediction and module identification. IEEE Journal of Biomedical and Health Informatics, 28, 1861–1871. https://doi.org/10.1109/JBHI.2024.3439713
Lin, R., Zhao, Z., Liu, Z., Kang, J., Zhang, K., Huang, X., ... & Yu, Y. (2026). Artificial intelligence in clinical oncology: Multimodal integration and translational development. Cancer Letters, 649, 218493. https://doi.org/10.1016/j.canlet.2026.218493
Liu, J., Yang, M., Bi, Y., Zhang, J., Yang, Y., Li, Y., Hong, S., Chen, K., & Li, X. (2025). Large language models enable tumor-type classification and localization of cancers of unknown primary from genomic data. Cell Reports Medicine, 6, 102332. https://doi.org/10.1016/j.xcrm.2025.102332
Lu, M. Y., Chen, B., Williamson, D. F. K., Chen, R. J., Liang, I., Ding, T., ... & Mahmood, F. (2024). A visual-language foundation model for computational pathology. Nature Medicine, 30(3), 863–874. https://doi.org/10.1038/s41591-024-02857-3
Moon, I., LoPiccolo, J., Baca, S. C., Sholl, L. M., Kehl, K. L., Hassett, M. J., ... & Gusev, A. (2023). Machine learning for genetics-based classification and treatment response prediction in cancer of unknown primary. Nature Medicine, 29, 2057–2067. https://doi.org/10.1038/s41591-023-02482-6
Ouhmouk, M., Baichoo, S., & Abik, M. (2025). Challenges in AI-driven multi-omics data analysis for oncology: Addressing dimensionality, sparsity, transparency and ethical considerations. Informatics in Medicine Unlocked, 57, 101679. https://doi.org/10.1016/j.imu.2025.101679
Qiu, Z., Kar, P., & Maulik, U. (2026). A genomic data analysis-based technique for personalized and precision medicine. Array, 30, 100965. https://doi.org/10.1016/j.array.2026.100965
Rescigno, P., & Greystoke, A. (2026). Perspectives on next-generation sequencing and artificial intelligence integration in oncology. Cancer Treatment and Research Communications, 46, 101054. https://doi.org/10.1016/j.ctarc.2026.101054
Roy, K. R., Moon, U. D., & Jamal, M. (2026). miRNA-mRNA Multi-omics Integration in Breast Cancer Staging. Advances in Biomarker Sciences and Technology., 9, 70–82. https://doi.org/10.1016/j.abst.2026.06.003
Sabit, H., Yadav, A. K., Salimy, S., Sakr, A., Abdel-Ghany, S., Wadan, A. S., Alqosaibi, A. I., Rashwan, R., AlGosaibi, Y. S., Alnamshan, M. M., Almulhim, J., Alaqeel, N. K., & Arneth, B. (2026). Multi-omics foundations in breast cancer and AI-driven advances: A clinical guide. Cancer Letters, 649, 218468. https://doi.org/10.1016/j.canlet.2026.218468
Sammut, S.-J., Crispin-Ortuzar, M., Chin, S.-F., Provenzano, E., Bardwell, H. A., Ma, W., ... & Caldas, C. (2022). Multi-omic machine learning predictor of breast cancer therapy response. Nature, 601, 623–629. https://doi.org/10.1038/s41586-021-04278-5
Schäffer, A. A., Chung, Y., Kammula, A. V., Ruppin, E., & Lee, J. S. (2024). A systematic analysis of the landscape of synthetic lethality-driven precision oncology. Med, 5(1), 73–89. https://doi.org/10.1016/j.medj.2023.12.009
Shah, B., Hussain, M., & Seth, A. (2025). Homologous recombination deficiency in ovarian and breast cancers: Biomarkers, diagnosis, and treatment. Current Issues in Molecular Biology, 47(8), 638. https://doi.org/10.3390/cimb47080638
Shanmugam, Y., & Ravikumar, L. (2026). Integrated multi-omics biomarker discovery workflow: Current advances and emerging clinical translation technologies. Advances in Biomarker Sciences and Technology, 8, 464–478. https://doi.org/10.1016/j.abst.2026.05.007
Su, Y., et al. (2024). Deep-ODX: Deep learning-based Oncotype DX recurrence score prediction from H&E whole-slide images. Expert Systems with Applications, 303, 130469. https://doi.org/10.1016/j.eswa.2025.130469
Wang, J., Zhu, H.-R., Xu, J., Fu, J., Liu, L.-Y., Chen, X.-Y., Chen, Z.-S., Lin, H.-W., & Gu, Z.-C. (2026). Oncology drug resistance prediction tools: Database infrastructure, algorithmic innovation, and clinical translation. Current Molecular Pharmacology, 19, 85–96. https://doi.org/10.2174/1874467226000103
Wang, T., & Ballester, P. J. (2021). MOGONET integrates multi-omics data using graph convolutional networks allowing patient classification and biomarker identification. Nature Communications, 12, 3445. https://doi.org/10.1038/s41467-021-23774-w
Webb, P. M., & Jordan, S. J. (2024). Global epidemiology of epithelial ovarian cancer. Nature Reviews Clinical Oncology, 21(5), 389–400. https://doi.org/10.1038/s41571-024-00881-3
Wolde, T., & Belay, M. (2026). Unraveling the complexity of gynecological cancers: challenges, resistance, and the road to precision medicine - A Narrative Review. Advances in Cancer Biology - Metastasis, 18, 100195. https://doi.org/10.1016/j.adcanc.2026.100195
Yin, J., Li, B., Xiong, H., Gan, J., & Liang, L. (2026). Methodological workflow combining single-cell RNA-seq and bulk transcriptomics to profile MDSCs in breast cancer. Translational Oncology, 63, 102605. https://doi.org/10.1016/j.tranon.2025.102605
Zhang, G., Peng, Z., Yan, C., Wang, J., Luo, J., & Luo, H. (2022). MultiGATAE: A novel cancer subtype identification method based on multi-omics and attention mechanism. Frontiers in Genetics, 13, 855629. https://doi.org/10.3389/fgene.2022.855629
Zhang, X., Xing, Y., Sun, K., & Guo, Y. (2021). OmiEmbed: A unified multi-task deep learning framework for multi-omics data. Cancers, 13(12), 3047. https://doi.org/10.3390/cancers13123047
Integrating Genomics, Transcriptomics, and Organoid-Based Drug Testing for Precision Cancer Care
Machine learning in cancer biology: transforming diagnosis, prognosis, and treatment in modern medical research
AI in Precision Oncology: Revolutionizing Early Diagnosis, Imaging and Targeted Therapies