A Normative Variational Autoencoder Scores Per-Cell Vulnerability in Parkinson’s Disease
Resumo
We recast per-cell vulnerability in Parkinson’s disease (PD) as deviation from a healthy norm—a transcriptome’s distance from healthy, not a prediction of which cells die. A variational autoencoder trained only on healthy midbrain nuclei gives each nucleus a deviation score, calibrated per cell type against a held-out healthy null. On the Smajić et al. (2022) cohort it separates PD from healthy nuclei (AUC = 0.623) with no disease label, concentrating in glia and vasculature; without marker genes, high-deviation glia carry the known reactive and disease-associated programs (ρ = 0.32, 0.30). The effect holds at donor level (exact permutation p = 0.002) and under an unseen-donor null, and reproduces in one external cohort but not a second.Referências
Fiorini, G., Li, Y., Fon, E. A., Farhan, S. M. K., Thomas, R. A., et al. (2025). Neural networks reveal novel gene signatures in parkinson disease from single-nuclei transcriptomes. npj Parkinson’s Disease, 11.
Habib, N., McCabe, C., Medina, S., Varshavsky, M., Kitsberg, D., et al. (2020). Disease-associated astrocytes in alzheimer’s disease and aging. Nature Neuroscience, 23(6):701–706.
Kamath, T., Abdulraouf, A., Burris, S. J., Langlieb, J., Gazestani, V., Nadaf, N. M., Balderrama, K., Vanderburg, C., and Macosko, E. Z. (2022). Single-cell genomic profiling of human dopamine neurons identifies a population that selectively degenerates in parkinson’s disease. Nature Neuroscience, 25(5):588–595.
Keren-Shaul, H., Spinrad, A., Weiner, A., Matcovitch-Natan, O., Dvir-Szternfeld, R., et al. (2017). A unique microglia type associated with restricting development of alzheimer’s disease. Cell, 169(7):1276–1290.
Kingma, D. P. and Welling, M. (2014). Auto-encoding variational bayes. International Conference on Learning Representations (ICLR).
Liddelow, S. A., Guttenplan, K. A., Clarke, L. E., Bennett, F. C., Bohlen, C. J., et al. (2017). Neurotoxic reactive astrocytes are induced by activated microglia. Nature, 541(7638):481–487.
Lopez, R., Regier, J., Cole, M. B., Jordan, M. I., and Yosef, N. (2018). Deep generative modeling for single-cell transcriptomics. Nature Methods, 15(12):1053–1058.
Lotfollahi, M., Naghipourfar, M., Luecken, M. D., et al. (2022). Mapping single-cell data to reference atlases by transfer learning. Nature Biotechnology, 40:121–130.
Mallick, K., Chakraborty, S., Mallik, S., and Bandyopadhyay, S. (2023). A scalable unsupervised learning of scrnaseq data detects rare cells through integration of structure-preserving embedding, clustering and outlier detection. Briefings in Bioinformatics, 24(3):bbad125.
Marquand, A. F., Kia, S. M., Zabihi, M., Wolfers, T., Buitelaar, J. K., and Beckmann, C. F. (2019). Conceptualizing mental disorders as deviations from normative functioning. Molecular Psychiatry, 24(10):1415–1424.
Marquand, A. F., Rezek, I., Buitelaar, J., and Beckmann, C. F. (2016). Understanding heterogeneity in clinical cohorts using normative models: beyond case-control studies. Biological Psychiatry, 80(7):552–561.
Martirosyan, A., Ansari, R., Pestana, F., Hebestreit, K., Gasparyan, H., et al. (2024). Unravelling cell type-specific responses to parkinson’s disease at single cell resolution. Molecular Neurodegeneration, 19(1):7.
Poulin, J.-F., Gaertner, Z., Moreno-Ramos, O. A., and Awatramani, R. (2020). Classification of midbrain dopamine neurons using single-cell gene expression profiling approaches. Trends in Neurosciences, 43(3):155–169.
Smajić, S., Prada-Medina, C. A., Landoulsi, Z., Ghelfi, J., Memet, S., et al. (2022). Single-cell sequencing of human midbrain reveals glial activation and a parkinson-specific neuronal state. Brain, 145(3):964–978.
Surmeier, D. J., Obeso, J. A., and Halliday, G. M. (2017). Selective neuronal vulnerability in parkinson disease. Nature Reviews Neuroscience, 18(2):101–113.
Wolf, F. A., Angerer, P., and Theis, F. J. (2018). Scanpy: large-scale single-cell gene expression data analysis. Genome Biology, 19(1):15.
Xu, K., Ding, Y., Hou, S., Zhan, W., Chen, N., Wang, J., and Sun, X. (2024). Domain adaptive and fine-grained anomaly detection for single-cell sequencing data and beyond. In Proceedings of the 33rd International Joint Conference on Artificial Intelligence (IJCAI).
Habib, N., McCabe, C., Medina, S., Varshavsky, M., Kitsberg, D., et al. (2020). Disease-associated astrocytes in alzheimer’s disease and aging. Nature Neuroscience, 23(6):701–706.
Kamath, T., Abdulraouf, A., Burris, S. J., Langlieb, J., Gazestani, V., Nadaf, N. M., Balderrama, K., Vanderburg, C., and Macosko, E. Z. (2022). Single-cell genomic profiling of human dopamine neurons identifies a population that selectively degenerates in parkinson’s disease. Nature Neuroscience, 25(5):588–595.
Keren-Shaul, H., Spinrad, A., Weiner, A., Matcovitch-Natan, O., Dvir-Szternfeld, R., et al. (2017). A unique microglia type associated with restricting development of alzheimer’s disease. Cell, 169(7):1276–1290.
Kingma, D. P. and Welling, M. (2014). Auto-encoding variational bayes. International Conference on Learning Representations (ICLR).
Liddelow, S. A., Guttenplan, K. A., Clarke, L. E., Bennett, F. C., Bohlen, C. J., et al. (2017). Neurotoxic reactive astrocytes are induced by activated microglia. Nature, 541(7638):481–487.
Lopez, R., Regier, J., Cole, M. B., Jordan, M. I., and Yosef, N. (2018). Deep generative modeling for single-cell transcriptomics. Nature Methods, 15(12):1053–1058.
Lotfollahi, M., Naghipourfar, M., Luecken, M. D., et al. (2022). Mapping single-cell data to reference atlases by transfer learning. Nature Biotechnology, 40:121–130.
Mallick, K., Chakraborty, S., Mallik, S., and Bandyopadhyay, S. (2023). A scalable unsupervised learning of scrnaseq data detects rare cells through integration of structure-preserving embedding, clustering and outlier detection. Briefings in Bioinformatics, 24(3):bbad125.
Marquand, A. F., Kia, S. M., Zabihi, M., Wolfers, T., Buitelaar, J. K., and Beckmann, C. F. (2019). Conceptualizing mental disorders as deviations from normative functioning. Molecular Psychiatry, 24(10):1415–1424.
Marquand, A. F., Rezek, I., Buitelaar, J., and Beckmann, C. F. (2016). Understanding heterogeneity in clinical cohorts using normative models: beyond case-control studies. Biological Psychiatry, 80(7):552–561.
Martirosyan, A., Ansari, R., Pestana, F., Hebestreit, K., Gasparyan, H., et al. (2024). Unravelling cell type-specific responses to parkinson’s disease at single cell resolution. Molecular Neurodegeneration, 19(1):7.
Poulin, J.-F., Gaertner, Z., Moreno-Ramos, O. A., and Awatramani, R. (2020). Classification of midbrain dopamine neurons using single-cell gene expression profiling approaches. Trends in Neurosciences, 43(3):155–169.
Smajić, S., Prada-Medina, C. A., Landoulsi, Z., Ghelfi, J., Memet, S., et al. (2022). Single-cell sequencing of human midbrain reveals glial activation and a parkinson-specific neuronal state. Brain, 145(3):964–978.
Surmeier, D. J., Obeso, J. A., and Halliday, G. M. (2017). Selective neuronal vulnerability in parkinson disease. Nature Reviews Neuroscience, 18(2):101–113.
Wolf, F. A., Angerer, P., and Theis, F. J. (2018). Scanpy: large-scale single-cell gene expression data analysis. Genome Biology, 19(1):15.
Xu, K., Ding, Y., Hou, S., Zhan, W., Chen, N., Wang, J., and Sun, X. (2024). Domain adaptive and fine-grained anomaly detection for single-cell sequencing data and beyond. In Proceedings of the 33rd International Joint Conference on Artificial Intelligence (IJCAI).
Publicado
19/10/2026
Como Citar
SILVA, Guilherme Felype Herculano da; MOURA, Karla Haryanna Santos; CASADO, Luana Dantas Pontes Espinola.
A Normative Variational Autoencoder Scores Per-Cell Vulnerability in Parkinson’s Disease. In: ENCONTRO NACIONAL DE INTELIGÊNCIA ARTIFICIAL E COMPUTACIONAL (ENIAC), 23. , 2026, Cuiabá/MT.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2026
.
p. 3250-3261.
ISSN 2763-9061.
DOI: https://doi.org/10.5753/eniac.2026.31299.
