Omics analysis for the Microglia project
This repository includes code and plots. Exploratory analysis and intermediate processing files are too large for this repository.
Full nominal and permuted eQTL and sQTL summary statistics per brain region are available from Zenodo at https://doi.org/10.5281/zenodo.4118605 (eQTL) and https://doi.org/10.5281/zenodo.4118403 (sQTL). Results for eQTL and sQTL meta-analysis (mashR and METASOFT) and colocalization (COLOC) are available from Zenodo at https://doi.org/10.5281/zenodo.4118676. Allelic information for all QTL analyses is available from Zenodo at https://doi.org/10.5281/zenodo.4301005.
Figure 1: Overview of the MiGA study.
1.Looking at covariates:
- Variance partition plots. Includes pairwise correlation plot, VP for 255 samples, VP by region and VP for European only (N=216).
2.Exploratory Analysis:
-
Exploratory plots after filters. General counts for age, sex, cause of death ... with 255 samples.
-
The PCAs are here.
-
Linear regression between the first 15 PCs and the covariates.
3.Expression of markers genes for 255 samples.
Differential expression analysis - PAIRWISE
Covariates: ~ sex + (1|donor_id) + age + tissue + (1|cause_of_death_categories) + C1 + C2 + C3 + C4 + picard_pct_mrna_bases + picard_summed_median + picard_pct_ribosomal_basesNumber of DE genes by comparison from Dream:
Filtered by: adj.P.Val<0.05 & AveExpr > 0 & abs(logFC) > 1| Comparison | Genes |
|---|---|
| MFG x STG | 6 |
| THA x SVZ | 387 |
| MFG x SVZ | 609 |
| THA x STG | 105 |
| THA x MFG | 119 |
| SVZ x STG | 909 |
Sex-related analysis
Our sex-related analysis = 0 genes at FDR 5%
Differential expression analysis for different DIAGNOSIS
Covariates: ~ main_diagnosis + sex + (1|donor_id) + age + tissue + (1|cause_of_death_categories) + C1 + C2 + C3 + C4 + picard_pct_mrna_bases + picard_summed_median + picard_pct_ribosomal_basesNumber of DE genes by comparison from Dream:
| Comparison | FDR 5% |
|---|---|
| Depression x CT | 0 |
| Dementia x CT | 24 |
| Schizophrenia + Bipolar x CT | 0 |
| Parkinson's x CT | 0 |
Age-related Analysis
Our age-related analysis using the tool Dream (N=255). removeBatchEffect to remove confounders from the expression table.
Effect size correlation plots.
DE genes with interaction term between age and region. Include scatter plots by region.
Age-related with PSI data.
Genotypes
Number of samples for downstream analysis. After GSA includes European only:
| Tissue | Before GSA filter | After GSA filter |
|---|---|---|
| MFG | 78 | 63 |
| STG | 64 | 55 |
| THA | 60 | 53 |
| SVZ | 56 | 45 |
| CC | 17 | 16 |
| CER | 7 | 4 |
| SN | 3 | 0 |
| Total | 285 samples | 236 samples |
Checking the numbers of samples from GSA. Click here.
Number of eGenes by brain region, including different number of PEER factors. Click here.
You can check the lists of eGenes, by brain region in this link.
Number of eGenes (includes covariate file with age, sex and 4 MDS ancestry):
| Tissue | Sample size | eGenes at qval 5% | eGenes at qval 10% |
|---|---|---|---|
| MFG | 63 | 199 | 267 |
| STG | 55 | 127 | 154 |
| THA | 53 | 148 | 196 |
| SVZ | 45 | 67 | 94 |
mashR pages - eQTL
The page with mashR results is here. Include a table with significative genes (lfsr < 0.10), heatmap of pairwise sharing by magnitude, metaplots and posterior.beta table.
Tissue specific effects: lfsr at 10%. Include lists of gene-SNP by region.
Tissue specific effects: lfsr at 5%. Include lists of gene-SNP by region.
mashR pages - sQTL
The page with mashR results is here. Includes a table with significative junction clusters (lfsr < 0.05) and a heatmap of pairwise sharing by magnitude.
Tissue specific effects: lfsr at 5%. Include lists of cluster-SNP by region.
Colocalization and fine-mapping
-
Qvalue sharing. Includes heatmap.
-
Fine Mapping COLOC PLAC-Seq.
-
Fine-mapping and COLOC overlaps.
Single-cell analysis
- Link for the single-cell results. Includes Quality Control and Downstream analysis.