The ongoing
COVID-۱۹ pandemic, driven by changing SARS-CoV-۲ variants, hashighlighted new concerns: the virus can affect the brain beyond the respiratory system. SARS-CoV-۲'sability to invade the nervous system through ACE۲ receptors raises worries about its impact onneurological diseases. We need to investigate if
COVID-۱۹ worsens cognitive decline and existingneurological conditions. Understanding how SARS-CoV-۲ enters the brain, causes inflammation, anddisrupts the blood-brain barrier is crucial for predicting and managing long-term neurological effects.Early and accurate diagnosis of neurodegenerative diseases (NDs) is crucial due to limited treatmentoptions and high costs associated with late or inaccurate diagnosis. Challenges include difficulties inmaking definitive diagnoses in early stages and predicting disease progression. Identifying bloodbiomarkers offers a promising avenue for easier, more cost-effective, and faster ND diagnosis,potentially leading to earlier intervention, better disease management, and improved patient outcomes.Real-time PCR is a standard method for gene expression measurement, limited in gene coverage.RNASeq analysis is used to identify gene expression differences but faces challenges in specific cellline analysis and requires large sample sizes. Single-cell RNA sequencing (scRNA-seq) offersunparalleled insights into cellular heterogeneity and gene expression at the individual cell level. Itenables the discovery of rare cell populations, identification of biomarkers, and understanding diseasemechanisms with unprecedented precision.This study aims to identify blood biomarkers for early diagnosis of neurodegenerative diseases inCOVID-۱۹ patients. To achieve this goal, scRNA-seq data related to COVID-۱۹, Alzheimer’s Disease(AD), and
Multiple Sclerosis (MS) were extracted from the GEO database, focusing on peripheral bloodmononuclear cells (PBMC). After analyzing scRNA-seq data and selecting genes with adjusted p-value< ۰.۰۵ and (logfc < -۰.۰۱ | logfc > ۰.۰۱), we employed a deep neural network, particularly an adversarialautoencoder with a classifier, to accurately classify normal and diseased samples based on their celltypes. Models were tailored to individual diseases and four blood tissue cell lines (monocytes, NK cells,B cells, and T cells). To evaluate the importance of genes in the classification process, we employedmultiplication of the weight matrix. Subsequently, we compared the results extracted from the modelsfor each disease and each cell line. These findings indicate that genes associated with
COVID-۱۹ maycontribute significantly to the development of neurological diseases such as AD and MS by affectinginflammatory and apoptotic pathways. This discovery holds promise for advancing drug developmentand early detection of post-coronavirus complications in individuals.