🧬 JCAP RNA-SEQ ANALYZER

An interactive R Shiny web app for RNA-seq analysis — integrating differential expression, power estimation, feature selection (Random Forest), dimensionality reduction, and pathway enrichment in one clean UI.

Built for bioinformatics teams, core facilities, and HPC environments.
No coding required. Fully Docker / Apptainer / Singularity compatible.

---

🚀 Features

• 📂 Upload expression matrix + phenotype file
• 🧬 Differential expression (limma-voom)
• 🌌 PCA + UMAP projections
• 🌋 Volcano plot
• 🌲 Random Forest + classifier metrics
• 📊 Functional enrichment (via enrichR)
• 📈 Power analysis & power curve
• 📉 ROC + AUC visualizations
• 📥 Downloadable results + logs
• 🧠 Error tracking via error_log.txt
• 🐳 Docker + Singularity + Apptainer support
• 📄 Embedded readme tab inside the app

---

---

🖼️ Screenshots

Differential Expression Results

Shows the top DEGs with Ensembl ID, gene symbol, description, p-value, logFC, and RF score.

---

Heatmap (Z-scored)

Clustered heatmap of top DEGs across all samples.

---

Pathway Enrichment Table

Enriched terms for selected genes (EnrichR).

---

Enrichment Barplot (Interactive)

Top terms visualized with -log10 adjusted P-values.

📁 Folder Structure

RNA_SEQ_APP/
├── app.R                     # Main Shiny app
├── run.sh                   # Docker run script
├── Dockerfile               # Docker image definition
├── Singularity.def          # Apptainer/Singularity container
├── Makefile                 # Build & run helper (Docker/SIF)
├── email_log.R              # Error email script
├── error_log.txt            # Runtime error log (auto-generated)
├── JCAP RNA_SEQ Readme.txt  # Embedded app readme (in-app)
└── www/
    └── Arcane_Alchemy_Theme.css


⚙️ Deployment Options
🐳 Run Locally with Docker
bash
Copy
Edit
git clone https://github.com/your-user/RNA_SEQ_APP.git
cd RNA_SEQ_APP
bash run.sh
Open your browser at:
http://localhost:8787

🧬 Run on HPC with Singularity or Apptainer
🔧 Build the container
bash
Copy
Edit
# Using Singularity
singularity build rna_seq_app.sif Singularity.def

# Or using Apptainer (drop-in replacement for Singularity)
apptainer build rna_seq_app.sif Singularity.def
🚀 Run the app
in bash

# Run with proper binding so the app can access code and CSS
singularity run --bind $(pwd):/app rna_seq_app.sif
# or
apptainer run --bind $(pwd):/app rna_seq_app.sif
📁 Make sure your app.R and www/ folder are in the current directory when using --bind.

🌐 Port Forwarding for Remote Access (e.g., HPC cluster)
If running on a remote server or cluster, enable port forwarding:

in bash

ssh -L 8080:localhost:8080 youruser@your-cluster
Then open your browser locally at:
http://localhost:8080

🧬 Run on HPC with SLURM + Singularity

🛠️ 1. Build Singularity Image (if not yet built)
b

in bash

singularity build rnaseq_shiny.sif Singularity.def
 🚀 2. Submit SLURM Job

in bash
sbatch submit_rna_seq_app.slurm

📍 3. Monitor Logs
 in bash

# Tail SLURM job logs
tail -f logs/slurm_<job_id>.out
tail -f logs/slurm_<job_id>.err
Replace <job_id> with your actual job number

🌐 4. Port Forwarding (Optional)

If running remotely, forward port 3838
in bash
ssh -L 3838:localhost:3838 youruser@hpc.cluster.edu

Then open: http://localhost:3838




🛠️ Using the Makefile
A Makefile is included to automate builds and runs:

Command	Description
make	Show help info
make docker	Build Docker image
make run	Run app using Docker (localhost:8787)
make run-dev	Run with local bind-mount for dev mode
make sif	Build Singularity image
make apptainer	Build Apptainer image
make run-sif	Run SIF interactively
make clean	Remove Docker image + SIF
📊 UI Tabs & Outputs
Tab	Description
Differential Expression	Top 50 DEGs + annotation + RF importance
PCA / UMAP Plots	Interactive projections of sample space
Volcano Plot	logFC vs -log10 P-adj with thresholds
Heatmap	Z-score clustered DE gene heatmap
Pathway Enrichment	Enrichr (All / Up / Down genes)
Power Summary / Curve	Power tables + curve from pwr package
Classification	RF model performance, predictions, ROC curve
Read Me	Rendered version of JCAP RNA_SEQ Readme.txt
🔁 Processing Workflow
t\
Counts + Phenotype
        ↓
Differential Expression (limma-voom)
        ↓
Top 50 DEGs → RF Feature Selection (Top 20)
        ↓
Classifier Training (Train/Test Split)
        ↓
Optional Enrichment (All / Up / Down)
        ↓
Dimensionality Reduction (PCA / UMAP)
        ↓
Visualization & Power Analysis

🧠 Tech Stack


R / Shiny

limma, edgeR, biomaRt, enrichR

randomForest, caret, pROC, pwr

umap, ggplot2, plotly

Docker, Singularity, Apptainer

Email alerts, Error logging, Makefile automation

📥 Downloadable Outputs

File	Description
DE_results_*.csv	Final DEGs with annotation + RF scores
enrichment_all.csv	All DEGs enrichment
enrichment_upregulated.csv	Upregulated gene enrichment
enrichment_downregulated.csv	Downregulated gene enrichment
rf_predictions.csv	Predictions table
rf_metrics.csv	Sensitivity, specificity, AUC
power_summary.csv	Power analysis summary

summary
📧 Error Monitoring (Optional)
Errors are auto-logged to error_log.txt.

You can schedule daily error emails via cron:

cron


📢 Feedback / Contributions
🐛 GitHub Issues
0 0 * * * Rscript /path/to/email_log.R

💡 Feature suggestions and PRs welcome

RNA-seq shouldn’t feel like black magic.
🧙‍♂️ JCAP RNA-SEQ ANALYZER turns your data into insights — fast, reproducibly, and interactively