This repository provides the code companion for the manuscript entitle: “Identification of p38 MAPK as a Major Therapeutic Target for Alzheimer’s Disease based on Integrative Pathway Activity Analysis and Validation in 3D Human Cellular Models.”
please contact pnaderiy [at] bidmc [dot] harvard [dot] edu for any
questions you may have about the codes.
First, you need to access processed RNA-Seq datasets from the ROSMAP, MSBB, and MAYO cohort via “Synape.org” (data accessions to be provided). These datasets require an approved Data Use Agreement to protect human subject privacy. You also need to download RNA-Seq datasets corresponding to the cellular models (data accession to be provided)
The RNA-Seq datasets and associated covariates should be placed into
respective directories under the folder preprocessed.
All codes relating to RNA-Seq data processing are placed in the codes
directory.
| Folder name | Description | Details |
|---|---|---|
codes |
Associated R-script | contains 4 subdirectories that organize codes for cleaning, pathway dysregulation analysis, plotting, and processing additional datasets. |
figures |
figure outputs. | Figures are saved in this directory according to their order in the manuscript |
Output |
Processed datasets | Dysregulated pathway profiles, pathway activity residuals, and similarity analysis results |
preprocessed |
Necessary data files | preprocessed RNA-Seq data from cellular models and human datasets. Background pathways and gene-level information are also provided to facilitate appropriate codes. |
tables |
output tables | dysregulated pathways, shared pathways across human cohorts and cell models, differentially expressed genes. |
reduced_pathway_output |
supplementary experiments | dysregulated pathways produced using a reduced background dataset for robustness analysis. |
| Folder name | file name | description |
|---|---|---|
00_new_msigdb |
01-MSigDBV6CONSERVATIVE.R |
Supplementary analysis: generating a reduced background pathway dataset for robustness analysis. |
00_new_msigdb |
02_MSigDB_XML.R |
Generating display friendly names for MSigDB Pathways |
01_cleaning_and_prep |
01-DEG_normalizer.R |
processing DEG files from multiple human cohorts to provide a concordant representation |
01_cleaning_and_prep |
02-clean_DEG_Tables.R |
processing DEG files from multiple cell lines to provide a concordant representation |
02_Dysregulated pathways |
01-FullPipeline_Mayo_ADvsControl.R |
Dysregulated pathway analysis in the Mayo cohort via PanomiR package |
02_Dysregulated pathways |
02-FullPipeline_MSBB_ADvsControl.R |
Dysregulated pathway analysis in the MSBB cohort via PanomiR package |
02_Dysregulated pathways |
03-FullPipeline_ROSMAP_ADvsControl.R |
Dysregulated pathway analysis in the Mayo cohort via PanomiR package |
02_Dysregulated pathways |
04-FullPipeline_A5vsG2B2.R |
Dysregulated pathway analysis in the A5 cells via PanomiR package |
02_Dysregulated pathways |
05-FullPipeline_D4vsG2B2.R |
Dysregulated pathway analysis in the D4 cells line via PanomiR package |
02_Dysregulated pathways |
06-FullPipeline_H105vsG2B2.R |
Dysregulated pathway analysis in the h10 cells via PanomiR package |
02_Dysregulated pathways |
07-FullPipeline_I47FvsG2B2.R |
Dysregulated pathway analysis in the I47F cells via PanomiR package |
02_Dysregulated pathways |
08-FullPipeline_I45FvsI47F.R |
Dysregulated pathway analysis in the I45F cells compared to I47F cells via PanomiR package |
02_Dysregulated pathways |
09-clean_Pathways.R |
cleaning up dysregulated pathway tables |
02_Dysregulated pathways |
10-Gene_based_correlation.R |
Correlation analysis of gene dysregulation across cell models and human cohorts |
02_Dysregulated pathways |
11-Pathway_based_correlation2_new.R |
Correlation analysis of gene dysregulation across cell models and human cohorts. Heatmap of similarity in Figure 3D |
03_plots |
01-assay_similarities_brains.R |
Venn diagrams, heatmaps, and correlation plots in Figure 2. Tables representing shared dysregulated pathways. Venn diagram in Figure 4. |
03_plots |
02-region_plotting.R |
Pathway activity heatmap in Figure 2 |
03_plots |
03-Maximal_Heatmap.R |
Pathway activity heatmap in Figure 5 |
03_plots |
04-Bar_plots.R |
Heatmaps of shared dysregulated pathways in Figure 4. Co-expression network plot |
03_plots |
05_PCA_covariates.R |
PCA-based determination of significant confounding variables in Figure S01. |
03_plots |
06-gene_correlation_similarities_2.R |
Correlation analysis of differentially expressed genes, Figure 2F |
03_plots |
07-get_ma_plots.R |
MA plots corresponding to Figure S02 |
03_plots |
09-assay_similarities_new.R |
Bar plots associated with Chi-squared tests in Figure 3C |
03_plots |
10_Batch_correction.R |
PCA plot in Figure S02 |
03_plots |
11_get_P38_Genes.R |
Activity of genes in the P38 MAPK Pathway in Figure 5A |
03_plots |
12_Visualize_gene_matrix.R |
Gene-based heatmap of similarity in Figure 3E |
03_plots |
S01-survival_brains.R |
P-value distribution analysis in Figure S02 |
external datasets |
01_external_dataset_clean.R |
cleaning iPSC gene expression data |
external datasets |
02_external_similarity_sq.R |
pathway-based similarity analysis Figure S03 |
external datasets |
03_External_heatmap.R |
pathway-based similarity analysis Figure S03 |
Supplementary files that are necessary for reproducing the study are
provided in the preprocssed/ directory
| File name | Description | Reference/Resource |
|---|---|---|
geneParameters.tsv |
GC-content and gene-length | ENSEMBL |
MSigDBPathGeneTab.RDS |
Background Pathways for the PanomiR package | MSigDB and PanomiR Packages |
MSigDBPathGeneTabLite.RDS |
Reduced overlap Background Pathways MSigDB, for robustness testing | MSigDB and PanomiR Packages |
MsigDB_jaccard.zip |
Jaccard overlap between MSigDB package, unzip before using | MSigDB |
MsigDB_display_names.csv |
clean, displayable names from MSigDB | MSigDB |
PCxN_MSigDB_withJaccard.RDS |
Pathway coexpression network of MSigDB database | MSigDB and PCxN |
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent