Network Medicine Approach for Analysis of Alzheimer's Disease Gene Expression Data

Plasma protein-based identification of neuroimage-driven subtypes in mild cognitive impairment via protein-protein interaction aware explainable graph propagational network

As an early indicator of dementia, mild cognitive impairment (MCI) requires specialized treatment according to its subtypes for the effective prevention and management of dementia progression. Based on the neuropathological characteristics, MCI can be classified into Alzheimer's disease (AD)-related cognitive impairment (ADCI) and subcortical vascular cognitive impairment (SVCI), being more likely to progress to AD and subcortical vascular dementia (SVD), respectively. For identifying MCI subtypes, plasma protein biomarkers are recently seen as promising tools due to their minimal invasiveness and cost-effectiveness in diagnostic procedures. Furthermore, the application of machine learning (ML) has led the preciseness in the biomarker discovery and the resulting diagnostics. Nevertheless, previous ML-based studies often fail to consider interactions between proteins, which are essential in complex neurodegenerative disorders such as MCI and dementia. Although protein-protein interactions (PPIs) have been employed in network models, these models frequently do not fully capture the diverse properties of PPIs due to their local awareness. This limitation increases the likelihood of overlooking critical components and amplifying the impact of noisy interactions. In this study, we introduce a new graph-based ML model for classifying MCI subtypes, called eXplainable Graph Propagational Network (XGPN). The proposed method extracts the globally interactive effects between proteins by propagating the independent effect of plasma proteins on the PPI network, and thereby, MCI subtypes are predicted by estimation of the risk effect of each protein. Moreover, the process of model training and the outcome of subtype classification are fully explainable due to the simplicity and transparency of XGPN's architecture. The experimental results indicated that the interactive effect between proteins significantly contributed to the distinct differences between MCI subtype groups, resulting in an enhanced classification performance with an average improvement of 10.0 % compared to existing methods, also identifying key biomarkers and their impact on ADCI and SVCI.

Identification of disease-specific bio-markers through network-based analysis of gene co-expression: A case study on Alzheimer's disease

Finding biomarker genes for complex diseases attracts persistent attention due to its application in clinics. In this paper, we propose a network-based method to obtain a set of biomarker genes. The key idea is to construct a gene co-expression network among sensitive genes and cluster the genes into different modules. For each module, we can identify its representative, i.e., the gene with the largest connectivity and the smallest average shortest path length to other genes within the module. We believe these representative genes could serve as a new set of potential biomarkers for diseases. As a typical example, we investigated Alzheimer's disease, obtaining a total of 16 potential representative genes, three of which belong to the non-transcriptome. A total of 11 out of these genes are found in literature from different perspectives and methods. The incipient groups were classified into two different subtypes using machine learning algorithms. We subjected the two subtypes to Gene Ontology analysis and Kyoto Encyclopedia of Genes and Genomes analysis with healthy groups and moderate groups, respectively. The two sub-type groups were involved in two different biological processes, demonstrating the validity of this approach. This method is disease-specific and independent; hence, it can be extended to classify other kinds of complex diseases.

Biomarker discovery process at binomial decision point (2BDP): Analytical pipeline to construct biomarker panel

A clinical incident is typically manifested by several molecular events; therefore, it seems logical that a successful diagnosis, prognosis, or stratification of a clinical landmark require multiple biomarkers. In this report, we presented a machine learning pipeline, namely "Biomarker discovery process at binomial decision point" (2BDP) that took an integrative approach in systematically curating independent variables (e.g., multiple molecular markers) to explain an output variable (e.g., clinical landmark) of binary in nature. In a logical sequence, 2BDP includes feature selection, unsupervised model development and cross validation. In the present work, the efficiency of 2BDP was demonstrated by finding three biomarker panels that independently explained three stages of Alzheimer's disease (AD) marked as Braak stages I, II and III, respectively. We designed three assortments from the entire cohort based on these Braak stages; subsequently, each assortment was split into two populations at Braak score I, II or III. 2BDP systematically integrated random forest and logistic regression fitting model to find biomarker panels with minimum features that explained these three assortments, e.g., significantly differentiated two populations segregated by Braak stage I, II or III, respectively. Thereafter, the efficacies of these panels were measured by the area under the curve (AUC) values of the receiver operating characteristic (ROC) plot. The AUC-ROC was calculated by two cross-validation methods. Final set of gene markers was a mix of novel and a priori established AD signatures. These markers were weighted by unique coefficients and linearly connected in a group of 2-10 to explain Braak stage I, II or III by AUC ≥ 0.8. Small sample size and a lack of distinctly recruited Training and Test sets were the limitations of the present undertaking; yet 2BDP demonstrated its capability to curate a panel of optimum numbers of biomarkers to describe the outcome variable with high efficacy.

Gene network inference from single-cell omics data and domain knowledge for constructing COVID-19-specific ICAM1-associated pathways

Introduction: Intercellular adhesion molecule 1 (ICAM-1) is a critical molecule responsible for interactions between cells. Previous studies have suggested that ICAM-1 triggers cell-to-cell transmission of HIV-1 or HTLV-1, that SARS-CoV-2 shares several features with these viruses via interactions between cells, and that SARS-CoV-2 cell-to-cell transmission is associated with COVID-19 severity. From these previous arguments, it is assumed that ICAM-1 can be related to SARS-CoV-2 cell-to-cell transmission in COVID-19 patients. Indeed, the time-dependent change of the ICAM-1 expression level has been detected in COVID-19 patients. However, signaling pathways that consist of ICAM-1 and other molecules interacting with ICAM-1 are not identified in COVID-19. For example, the current COVID-19 Disease Map has no entry for those pathways. Therefore, discovering unknown ICAM1-associated pathways will be indispensable for clarifying the mechanism of COVID-19. Materials and methods: This study builds ICAM1-associated pathways by gene network inference from single-cell omics data and multiple knowledge bases. First, single-cell omics data analysis extracts coexpressed genes with significant differences in expression levels with spurious correlations removed. Second, knowledge bases validate the models. Finally, mapping the models onto existing pathways identifies new ICAM1-associated pathways. Results: Comparison of the obtained pathways between different cell types and time points reproduces the known pathways and indicates the following two unknown pathways: (1) upstream pathway that includes proteins in the non-canonical NF-κB pathway and (2) downstream pathway that contains integrins and cytoskeleton or motor proteins for cell transformation. Discussion: In this way, data-driven and knowledge-based approaches are integrated into gene network inference for ICAM1-associated pathway construction. The results can contribute to repairing and completing the COVID-19 Disease Map, thereby improving our understanding of the mechanism of COVID-19.

Longitudinal APOE4- and amyloid-dependent changes in the blood transcriptome in cognitively intact older adults

BACKGROUND

Gene expression is dysregulated in Alzheimer's disease (AD) patients, both in peripheral blood and post mortem brain. We investigated peripheral whole-blood gene (co)expression to determine molecular changes prior to symptom onset.

METHODS

RNA was extracted and sequenced for 65 cognitively healthy F-PACK participants (65 (56-80) years, 34 APOE4 non-carriers, 31 APOE4 carriers), at baseline and follow-up (interval: 5.0 (3.4-8.6) years). Participants received amyloid PET at both time points and amyloid rate of change derived. Accumulators were defined with rate of change ≥ 2.19 Centiloids. We performed differential gene expression and weighted gene co-expression network analysis to identify differentially expressed genes and networks of co-expressed genes, respectively, with respect to traits of interest (APOE4 status, amyloid accumulation (binary/continuous)), and amyloid positivity status, followed by Gene Ontology annotation.

RESULTS

There were 166 significant differentially expressed genes at follow-up compared to baseline in APOE4 carriers only, whereas 12 significant differentially expressed genes were found only in APOE4 non-carriers, over time. Among the significant genes in APOE4 carriers, several had strong evidence for a pathogenic role in AD based on direct association scores generated from the DISQOVER platform: NGRN, IGF2, GMPR, CLDN5, SMIM24. Top enrichment terms showed upregulated mitochondrial and metabolic pathways, and an exacerbated upregulation of ribosomal pathways in APOE4 carriers compared to non-carriers. Similarly, there were 33 unique significant differentially expressed genes at follow-up compared to baseline in individuals classified as amyloid negative at baseline and positive at follow-up or amyloid positive at both time points and 32 unique significant differentially expressed genes over time in individuals amyloid negative at both time points. Among the significant genes in the first group, the top five with the highest direct association scores were as follows: RPL17-C18orf32, HSP90AA1, MBP, SIRPB1, and GRINA. Top enrichment terms included upregulated metabolism and focal adhesion pathways. Baseline and follow-up gene co-expression networks were separately built. Seventeen baseline co-expression modules were derived, with one significantly negatively associated with amyloid accumulator status (r² =  - 0.25, p = 0.046). This was enriched for proteasomal protein catabolic process and myeloid cell development. Thirty-two follow-up modules were derived, with two significantly associated with APOE4 status: one downregulated (r² =  - 0.27, p = 0.035) and one upregulated (r² = 0.26, p = 0.039) module. Top enrichment processes for the downregulated module included proteasomal protein catabolic process and myeloid cell homeostasis. Top enrichment processes for the upregulated module included cytoplasmic translation and rRNA processing.

CONCLUSIONS

We show that there are longitudinal gene expression changes that implicate a disrupted immune system, protein removal, and metabolism in cognitively intact individuals who carry APOE4 or who accumulate in cortical amyloid. This provides insight into the pathophysiology of AD, whilst providing novel targets for drug and therapeutic development.

Shared genetic loci between Alzheimer's disease and multiple sclerosis: Crossroads between neurodegeneration and immune system

BACKGROUND

Neuroinflammation is involved in the pathophysiology of Alzheimer's disease (AD), including immune-linked genetic variants and molecular pathways, microglia and astrocytes. Multiple Sclerosis (MS) is a chronic, immune-mediated disease with genetic and environmental risk factors and neuropathological features. There are clinical and pathobiological similarities between AD and MS. Here, we investigated shared genetic susceptibility between AD and MS to identify putative pathological mechanisms shared between neurodegeneration and the immune system.

METHODS

We analysed GWAS data for late-onset AD (N cases = 64,549, N controls = 634,442) and MS (N cases = 14,802, N controls = 26,703). Gaussian causal mixture modelling (MiXeR) was applied to characterise the genetic architecture and overlap between AD and MS. Local genetic correlation was investigated with Local Analysis of [co]Variant Association (LAVA). The conjunctional false discovery rate (conjFDR) framework was used to identify the specific shared genetic loci, for which functional annotation was conducted with FUMA and Open Targets.

RESULTS

MiXeR analysis showed comparable polygenicities for AD and MS (approximately 1800 trait-influencing variants) and genetic overlap with 20% of shared trait-influencing variants despite negligible genetic correlation (rg = 0.03), suggesting mixed directions of genetic effects across shared variants. conjFDR analysis identified 16 shared genetic loci, with 8 having concordant direction of effects in AD and MS. Annotated genes in shared loci were enriched in molecular signalling pathways involved in inflammation and the structural organisation of neurons.

CONCLUSIONS

Despite low global genetic correlation, the current results provide evidence for polygenic overlap between AD and MS. The shared loci between AD and MS were enriched in pathways involved in inflammation and neurodegeneration, highlighting new opportunities for future investigation.

Alterations in Immune-Related Defensin Alpha 4 (DEFA4) Gene Expression in Health and Disease

Defensin Alpha 4 (DEFA4) is the fourth member of the Alpha Defensins family known as a part of antimicrobial peptides in the innate immune system. DEFA4 has a strong preference to kill Gram-negative bacteria more than Gram-positive bacteria. In addition, DEFA4 exhibits antiviral activity against human immunodeficiency virus type 1 (HIV-1) in vitro. Moreover, DEFA4 can act as an inhibitor of corticosterone production (Corticostatin). On the other hand, alternations in DEFA4 gene expression have been reported in different disorders such as diseases related to inflammation and immunity dysfunction, brain-related disorders, and various cancers. The up-regulation of DEFA4 appears to be involved in the malignant transformation or aggressive form of cancer. Interestingly, the modified version of DEFA4 fragment (1-11) was potent and efficient against antibiotic-resistant bacteria. This review provides a general background abSaudi Arabia out DEFA4 and sheds light on changes in DEFA4 gene expression in different diseases. The paper also discusses other aspects related to DEFA4 as an antimicrobial and antiviral agent. The research was conducted based on available articles obtained from databases starting from 1988 to the present.

Putative Factors Interfering Cell Cycle Re-Entry in Alzheimer's Disease: An Omics Study with Differential Expression Meta-Analytics and Co-Expression Profiling

BACKGROUND

Neuronal cell cycle re-entry (CCR) is a mechanism, along with amyloid-β (Aβ) oligomers and hyperphosphorylated tau proteins, contributing to toxicity in Alzheimer's disease (AD).

OBJECTIVE

This study aimed to examine the putative factors in CCR based on evidence corroboration by combining meta-analysis and co-expression analysis of omic data.

METHODS

The differentially expressed genes (DEGs) and CCR-related modules were obtained through the differential analysis and co-expression of transcriptomic data, respectively. Differentially expressed microRNAs (DEmiRNAs) were extracted from the differential miRNA expression studies. The dysregulations of DEGs and DEmiRNAs as binary outcomes were independently analyzed by meta-analysis based on a random-effects model. The CCR-related modules were mapped to human protein-protein interaction databases to construct a network. The importance score of each node within the network was determined by the PageRank algorithm, and nodes that fit the pre-defined criteria were treated as putative CCR-related factors.

RESULTS

The meta-analysis identified 18,261 DEGs and 36 DEmiRNAs, including genes in the ubiquitination proteasome system, mitochondrial homeostasis, and CCR, and miRNAs associated with AD pathologies. The co-expression analysis identified 156 CCR-related modules to construct a protein-protein interaction network. Five genes, UBC, ESR1, EGFR, CUL3, and KRAS, were selected as putative CCR-related factors. Their functions suggested that the combined effects of cellular dyshomeostasis and receptors mediating Aβ toxicity from impaired ubiquitination proteasome system are involved in CCR.

CONCLUSION

This study identified five genes as putative factors and revealed the significance of cellular dyshomeostasis in the CCR of AD.

LncRNA Xist, X-chromosome Instability and Alzheimer's Disease

Neurodegenerative Diseases (NDD) are the major contributors to age-related causes of mental disability on a global scale. Most NDDs, like Alzheimer's Disease (AD), are complex in nature - implying that they are multi-parametric both in terms of heterogeneous clinical outcomes and underlying molecular paradigms. Emerging evidence from high throughput genomic, transcriptomic and small RNA sequencing experiments hint at the roles of long non-coding RNAs (lncRNAs) in AD. X-inactive Specific Transcript (XIST), a component of the Xic, the X-chromosome inactivation centre, is an RNA gene on the X chromosome of the placental mammals indispensable for the X inactivation process. An extensive literature survey shows that aberrations in Xist expression and in some cases, a disruption of the Xchromosome inactivation as a whole play a significant role in AD. Considering the enormous potential of Xist as an endogenous silencing molecule, the idea of using Xist as a non-conventional chromosome silencer to treat diseases harboring chromosomal alterations is also being implemented. Comprehensive knowledge about how Xist could play such a role in AD is still elusive. In this review, we have collated the available knowledge on the possible Xist involvement and deregulation from the perspective of molecular mechanisms governing NDDs with a primary focus on Alzheimer's disease. Possibilities of XIST mediated therapeutic intervention and linkages between XIC and preferential predisposition of females to AD have also been discussed.