Schizophrenia Transcriptomic and ceRNA Network Analyses

This repository contains R scripts used for transcriptomic and competing endogenous RNA (ceRNA) regulatory network analyses in schizophrenia across multiple tissues.

All analyses were conducted using publicly available datasets. No individual-level or restricted-access data are included.

---

📄 Related Publications

The analyses implemented in this repository support the following peer-reviewed publications:

• Scientific Reports (2021)
  Bioinformatics analysis of long non-coding RNA-associated competing endogenous RNA network in schizophrenia
  DOI: 10.1038/s41598-021-03993-3

• Scientific Reports (2021)
  Long non-coding RNA-associated competing endogenous RNA axes in the olfactory epithelium in schizophrenia: a bioinformatics analysis
  DOI: 10.1038/s41598-021-04326-0

• Frontiers in Psychiatry (2022)
  Identification of key long non-coding RNA-associated competing endogenous RNA axes in Brodmann Area 10 brain region of schizophrenia patients
  DOI: 10.3389/fpsyt.2022.1010977

---

🧬 Overview

The repository includes scripts for:

• Microarray preprocessing and normalization
• Quality control and exploratory analyses (including PCA)
• Differential gene expression analysis using limma
• Integration of curated RNA-RNA interaction databases
• Construction and analysis of ceRNA regulatory networks
• Tissue-specific transcriptomic analyses

---

📁 Repository Structure

SCZ-ceRNA-analysis/
├── README.md
├── codes/
│   ├── 01_HPC_BA46_STR_codes.R
│   ├── 02_LB_codes.R
│   ├── 03_OE_codes.R
│   └── 04_BA10_codes.R
└── papers/
    ├── fpsyt-13-1010977.pdf
    ├── s41598-021-03993-3.pdf
    └── s41598-021-04326-0.pdf

---

🧪 Methods Summary

Transcriptomic analyses were conducted using publicly available microarray datasets. Raw intensity files were preprocessed and normalized using the Robust Multi-array Average (RMA) method. Quality control and exploratory analyses, including principal component analysis (PCA), were performed to assess sample-level variation and identify potential outliers.

Differential gene expression analysis was carried out using linear models implemented in the limma package, with empirical Bayes moderation applied to improve variance estimation. Multiple testing correction was performed using the Benjamini–Hochberg false discovery rate (FDR) approach.

Differentially expressed genes were identified based on statistical significance and effect size thresholds. Experimentally supported RNA–RNA interaction databases were integrated to infer miRNA–mRNA and lncRNA–miRNA interactions. These interactions were combined to construct ceRNA regulatory networks.

Correlation analyses were performed to support ceRNA relationships and to evaluate expression concordance between regulatory RNA components. All analyses were conducted in R using Bioconductor and CRAN packages, following the methodological descriptions reported in the associated publications.

---

⚠️ Notes on Reproducibility

• Scripts were developed using local file paths and dataset-specific inputs.
• Public datasets must be downloaded separately and file paths updated accordingly.