Search

Myelin is an essential component of the nervous system and myelin damage causes demyelination diseases. Myelin is a sheet of oligodendrocyte membrane wrapped around the neuronal axon. In the fluorescent images, experts manually identify myelin by co-localization of oligodendrocyte and axonal membranes that fit certain shape and size criteria. Because myelin wriggles along x-y-z axes, machine learning is ideal for its segmentation. However, machine-learning methods, especially convolutional neural networks (CNNs), require a high number of annotated images, which necessitate expert labor. To facilitate myelin annotation, we developed a workflow and software for myelin ground truth extraction from multi-spectral fluorescent images. Additionally, to the best of our knowledge, for the first time, a set of annotated myelin ground truths for machine learning applications were shared with the community., Competing Interests: No competing interests were disclosed., (Copyright: © 2023 Çapar A et al.)

Objective: Myelin is essential for a healthy nervous system. Myelin formed by oligodendrocytes, accelerates impulse propagation and supports neuronal survival. Demyelination leads to neurodegeneration. In multiple sclerosis (MS) immune attack results in demyelination. Our goal is to dissect interactions among oligodendrocytes, neurons, and immune cells to improve our understanding of myelination and the demyelinating diseases. We aim to discover new targets for remyelination therapies. Methods: We are building tools for analyzing protein-protein and cell-to-cell interactions. To identify genes involved in myelination and MS, we developed a bioinformatics-based strategy, interactome analysis, which combines proteome and gene expression methodologies. Identified genes are evaluated in peripheral blood of MS patients and in mouse models. To accelerate drug discovery, 23 machine learning-based methodologies were assessed for myelin identification in fluorescent microscopy images. Results: Interactome analysis identified hundreds of proteinprotein interactions between pairs of oligodendrocytes, neurons, macrophages, microglia, and T cells. Most significant interactions are further analyzed in vivo and in vitro. Our customizedconvolutional neural network and Boosted Tree methods segmented myelin at over 98% accuracy. Identified myelin can be quantified and visualized in 3D. Conclusion: The interactome analysis yielded novel genes that are likely to be linked to MS. Machine learning-based methodologies are very effective in accelerating myelin quantification and thus, drug screens against demyelinating diseases such as MS. Overall, our innovative analysis strategies employing computer assistance produced novel avenues of exploration for myelination and demyelinating diseases. This study was supported TUBITAK (218S495,316S026), Istanbul Medipol University (BAP2018/06), and Turkish Academy of Sciences (GEBIP). [ABSTRACT FROM AUTHOR]