Nathan Sahelijo

Mr. Nathan Sahelijo


Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
22 December 2022 in Molecular Psychiatry
Verified
Journal Article
Closed access
PDF file
RSRichard Sherva
Richard Sherva
RZRui Zhang
Rui Zhang
NSNathan Sahelijo
Nathan Sahelijo
GJGyungah Jun
Gyungah Jun
TATori Anglin
Tori Anglin
MLMark W. Logue
Mark W. Logue

29 February 2024 in Translational Psychiatry
Verified
Journal Article
Open access
PDF file
RPRebecca Panitch
Rebecca Panitch
NSNathan Sahelijo
Nathan Sahelijo
JHJunming Hu
Junming Hu
KNKwangsik Nho
Kwangsik Nho
DBDavid A. Bennett
David A. Bennett
GJGyungah Jun
Gyungah Jun
The joint effects of APOE genotype and DNA methylation on Alzheimer disease (AD) risk is relatively unknown. We conducted genome-wide methylation analyses using 2,021 samples in blood (91 AD cases, 329 mild cognitive impairment, 1,391 controls) and 697 samples in brain (417 AD cases, 280 controls). We identified differentially methylated levels in AD compared to controls in an APOE genotype-specific manner at 25 cytosine-phosphate-guanine (CpG) sites in brain and 36 CpG sites in blood. Additionally, we identified seven CpG sites in the APOE region containing TOMM40, APOE, and APOC1 genes with P < 5 × 10−8 between APOE ε4 carriers and non-carriers in brain or blood. In brain, the most significant CpG site hypomethylated in ε4 carriers compared to non-carriers was from the TOMM40 in the total sample, while most of the evidence was derived from AD cases. However, the CpG site was not significantly modulating expression of these three genes in brain. Three CpG sites from the APOE were hypermethylated in APOE ε4 carriers in brain or blood compared in ε4 non-carriers and nominally significant with APOE expression in brain. Three CpG sites from the APOC1 were hypermethylated in blood, which one of the 3 CpG sites significantly lowered APOC1 expression in blood using all subjects or ε4 non-carriers. Co-methylation network analysis in blood and brain detected eight methylation networks associated with AD and APOE ε4 status. Five of the eight networks included genes containing network CpGs that were significantly enriched for estradiol perturbation, where four of the five networks were enriched for the estrogen response pathway. Our findings provide further evidence of the role of APOE genotype on methylation levels associated with AD, especially linked to estrogen response pathway.

1 December 2024 in Alzheimer's & Dementia
Verified
Journal Article
Open access
DPDhawal Priyadarshi
Dhawal Priyadarshi
HHHeng Huang
Heng Huang
NSNathan Sahelijo
Nathan Sahelijo
RSReza Shirkavand
Reza Shirkavand
GJGyungah R Jun
Gyungah R Jun
Background: The prohibitive costs of drug development for Alzheimer’s Disease (AD) emphasize the need for alternative in silico drug repositioning strategies. Graph learning algorithms, capable of learning intrinsic features from complex network structures, can leverage existing databases of biological interactions to improve predictions in drug efficacy. We developed a novel machine learning framework, the PreSiBOGNN, that integrates muti‐modal information to predict cognitive improvement at the subject level for precision medicine in AD.Method: The graph neural network framework integrates four layers of input data including transcriptome, proteome, drug, and subject to connect to bipartite graphs in the United Kingdom Biobank (UKBB). Medication usage, clinical, and GWAS data were downloaded for 48187 subjects with first and second cognitive exams from the UKBB. The protein and transcriptome layers were constructed using the String database and gene coexpression networks generated from single nuclei RNA data (Sahelijo et al. 2022). Layers were connected by binary bipartite graphs constructed using drug information from the UniProt and DrugCentral databases. Sequential Graph Attention Networks convoluted embedded features generated by each layer in a hierarchical order: 1. gene‐gene, 2. gene‐protein, 3. protein‐protein, 4. protein‐drug, 5. drug‐subj. Feature embeddings were decoded using a multilayer perceptron to predict cognitive improvement between the first and second cognitive exams. Two models compared methods of data aggregation. The first model follows a strict hierarchy, aggregating inter‐layer data in a single direction (gene > protein > drug > subj). The second model allows for fluid message passing between inter‐layers. We used 60% of the UKBB subjects for training and 30% for validation. We assessed the model with the greatest training accuracy using the remaining 10% of subjects.Result: We observed that the performance of the strict message‐passing model attained validation and test accuracy of 57.7% and 51.2%, respectively. Performance of the fluid message passing model improved prediction accuracy with 61.3% and 58.1% in the test set.Conclusion: Our investigation suggests the feasibility of the PreSiBOGNN framework to infer cognitive improvement of existing drugs by integrating medication, multi‐omics, and clinical data. Future work will focus on model optimizations and the integration of additional modalities including compound‐specific fingerprint data.


25 December 2025 in Alzheimer's & Dementia
Verified
Journal Article
Open access
NSNathan Sahelijo
Nathan Sahelijo
DGDaniel Goldstein
Daniel Goldstein
GJGyungah R Jun
Gyungah R Jun
Background: Genetic subtyping can identify subpopulations at risk for developing Alzheimer Disease (AD), but it cannot detect definitive disease progression. Plasma biomarkers enable a cost‐effective, noninvasive approach to early detection of AD disease. Method: We defined genetic subtypes by calculating cell‐based polygenic risk scores (cbPRSs) with a suggestive significance ( p < 1×10 ‐5 ) in 8,481subjects from the Framingham Heart Study (FHS) using previously described procedures (Sahelijo et al., 2023). We defined high‐ and low‐risk individuals for four cell‐specific subtypes using the first and fourth quartiles of the cbPRS distribution in FHS. We analyzed differential expression of 1377 SOMAscan aptamer‐based plasma proteins between high‐ and low‐risk individuals in FHS using the LIMMA R software package, adjusting for age and sex. Pathway analysis was conducted for nominally significant differentially expressed proteins (DEP) with p <0.05 using Metascape software. Finally, we validated the identified subtype‐specific DEPs using clinically or neuropathologically defined AD diagnosis in autopsied brains from Boston University Alzheimer's Disease Research Center to identify the overlap of the subtype DEP with AD DEP using gene set enrichment analysis (GSEA). Result: Genetic subtypes were defined for two astrocyte and two oligodendrocyte networks with low and high‐risk sample sizes ranging from 2,072 to 2,090 subjects in FHS. Differential expression analysis revealed 59 to 91 nominally significant proteins with some overlapping proteins among the subtypes. All four subtypes exhibited enrichment for pathways associated with cytokine regulation and immune response. Top plasma DEP for Ast‐M2 was the MMP‐8 protein, which was validated in brain DEP for clinical and pathological AD (p AstM2 = 4.63×10 ‐5 , p clinicalAD =9.98×10 ‐5 , p PathAD =3.2×10 ‐3 ). Moreover, using GSEA we identified significant enrichment of DEP in clinical AD with Ast‐M2 DEP (p AstM2 =1.30×10 ‐2 ) and pathological AD with Ast‐M10 DEP (p AstM10 =3.72×10 ‐4 ). Conclusion: Plasma protein levels from at‐risk individuals for cell‐based genetic subtypes exhibited considerable enrichment for neuroinflammatory proteins, indicating potential as genetic‐subtype‐specific biomarkers. These results highlight plasma biomarkers as cost‐effective tools for early detection in genetically at‐risk individuals of specific brain cell types.

10 April 2026 in Journal of Clinical Medicine
Verified
Journal Article
Open access
PDF file
DGDaniel Goldstein
Daniel Goldstein
NSNathan Sahelijo
Nathan Sahelijo
DPDhawal Priyadarshi
Dhawal Priyadarshi
RPRebecca Panitch
Rebecca Panitch
KNKwangsik T. Nho
Kwangsik T. Nho
GJGyungah Jun
Gyungah Jun
Background: Polygenic risk scores for Alzheimer’s disease (AD), organized by gene networks shared between the blood and brain, may provide insights into underlying disease mechanisms common to both tissues. Methods: We derived a blood–brain network-based polygenic risk score (nbPRS) from AD-associated genetic variants for three blood-brain networks, selected by the preservation of blood and brain gene co-expression networks, and AD association. Participants from the Alzheimer’s Disease Neuroimaging Initiative (ADNI, n = 1109), Framingham Heart Study (FHS, n = 8310), the Religious Orders Study Memory Aging Project (ROSMAP, n = 1215), and Mount Sinai Brain Bank (MSBB, n = 323) were stratified into low- and high-nbPRS subgroups, then profiled using longitudinal and cross-sectional data. We compared the conversion from normal cognition to AD between nbPRS subgroups. Genes differentially expressed among low- and high-nbPRS individuals were profiled with classical neuropathological markers and we investigated potential biologically relevant pathways for the genes significantly expressed in high-risk individuals. Results: Individuals with high nbPRS in three AD-associated networks (M2, M6, M14) demonstrated significant impairment in executive function and memory performance, whereas high-risk individuals in networks M2 and M14 had significantly reduced hippocampal volume. We observed high-risk individuals in M2 and M14 developed AD at twice the rate of low-risk individuals in these networks. HLA genes were differentially expressed with transcriptome-wide significance among low- and high-nbPRS individuals in M14 and associated with neuroinflammatory and tau pathology. Conclusions: Polygenic risk scores derived from blood and brain networks can differentiate individuals with a high risk of AD conversion.