🧬 Cell-Surface glycoRNA-Based Cancer Detection

         

---

📌 Project Summary

This project focuses on the development of chip-based cancer detection via surface glycoRNA markers derived from both cancer cell lines and extracellular vesicles (EVs).

• 📡 Data Source: Raw ncRNA sequencing data from CD Genomics, USA
• 🧪 Institute: Australian Institute for Bioengineering and Nanotechnology (AIBN), University of Queensland (UQ)
• 🔍 Objective: Discover and validate glycoRNA markers for cancer detection using RNA-seq, bioinformatics pipelines, and machine learning models

---

🔬 Workflow Overview

🔄 RNA-Seq Processing Pipeline

# Step 1: Quality Control
fastqc *.fastq.gz

# Step 2: Trimming
trimmomatic PE input_1.fastq.gz input_2.fastq.gz ...

# Step 3: Alignment
STAR --runThreadN 8 --genomeDir path/to/genome --readFilesIn input_1.fastq.gz input_2.fastq.gz

# Step 4: Counting
featureCounts -a annotation.gtf -o counts.txt aligned_reads.bam

📦 R Packages

library(DESeq2)
library(edgeR)
library(ggplot2)
library(pheatmap)
library(clusterProfiler)
library(org.Hs.eg.db)
library(ncRNAtools)
library(RNAcentral)

---

📚 Project Chapters

📁 CHAPTER 01: Cancer Cell Line Analysis

Comparative Analysis (ADMSC & BMMSC vs Cancer Cell Lines)

🔬 Cancer Cell Lines:

Cancer Type	Cell Line(s)
Cervical	HeLa
Breast	MCF7 (Epithelial), MDA-MB-231 (Mesenchymal)
Lung	A549, H1975
Controls	ADMSC, BMMSC

📊 Analysis Performed:

1. Volcano Plots

   • Annotated with Y-RNAs and U-RNAs
   • Comparisons:
     • ADMSC vs HeLa
     • ADMSC vs MCF7
     • ADMSC vs MDA-MB-231
     • BMMSC vs HeLa
     • BMMSC vs MCF7
     • BMMSC vs MDA-MB-231
     • ADMSC vs A549
     • ADMSC vs H1975
     • BMMSC vs A549
     • BMMSC vs H1975

2. Heatmaps

   • All Y-RNAs and U-RNAs across all cell lines
   • Metrics: Log2 Fold Change, Z-score Expression

3. Correlation Plots

   • Expression correlation across Y-RNA signatures

---

📁 CHAPTER 02: Extracellular Vesicles (EVs)

🔬 Focus: Identification of glycoRNA markers in EVs secreted by cancer cell lines.

• Extracted EV-specific DEGs
• Annotated with ncRNAtools and RNAcentral
• Identified diagnostic glycoRNAs shared across multiple cancer types

📌 EVs highlighted as a promising source of non-invasive biomarkers

---

📁 CHAPTER 03: Metastasis & EMT Transition

🧬 Epithelial–Mesenchymal Transition (EMT)
Focuses on progression from primary tumor to metastatic state in breast cancer.

🧫 Cell Lines:

• MCF7 – Epithelial breast cancer cell line
• MDA-MB-231 – Mesenchymal/metastatic breast cancer cell line
• TM6 – Treated MCF7 (6-day treatment, induced EMT)

🧪 Comparison:

1. Y-RNA and U-RNA in:

   • TM6 vs ADMSC
   • TM6 vs BMMSC

2. Marker Signature Comparisons:

   • MDA-MB-231 vs MCF7
   • TM6 vs MCF7
   • TM6 vs MDA-MB-231

3. Correlation Analysis:

   • TM6 vs MDA-MB-231 to validate metastatic similarity

🚨 EMT is a critical transition step in cancer metastasis, and glycoRNAs here serve as potential metastasis indicators.

---

🤖 Machine Learning Pipeline

🧠 Goals:

• Identify cancer-type specific markers
• Detect common markers across all cancer types
• Distinguish metastatic (EMT) from non-metastatic profiles

📈 Framework:

• Models: Random Forest, SVM, Logistic Regression
• Features: DE glycoRNAs (Y-RNA, U-RNA)
• Training: Stratified 5-Fold Cross-Validation
• Evaluation Metrics: AUROC, AUPRC, Accuracy, F1-score

🧬 Output:

• Ranked glycoRNA markers per cancer type
• Heatmaps, PR/ROC Curves, Confusion Matrices

---

🧪 qRT-PCR Validation

• Designed primers for top-ranked markers (Y-RNA, U-RNA)
• Validated expression in patient tissue samples
• Correlated with model-predicted marker rankings

---

📎 Tools & Technologies

Category	Tools/Packages
Sequencing QC	FastQC, Trimmomatic
Alignment	STAR
Quantification	featureCounts
DE Analysis	DESeq2, edgeR
ML Modeling	scikit-learn, caret (R), custom pipelines
Annotation	ncRNAtools, RNAcentral
Visualization	ggplot2, pheatmap, EnhancedVolcano

---

📂 Folder Structure

2022_MPXV_Project/
│
├── data/                  # FASTQ, BAM, and Count files
├── results/
│   ├── chapter1/          # Cell-line analysis outputs
│   ├── chapter2/          # EV-specific markers
│   ├── chapter3/          # EMT and metastasis analysis
│
├── figure/
│   └── git_readme/        # Logos (AIBN, CD Genomics, UQ)
│
├── scripts/               # All preprocessing, analysis, ML scripts
├── qPCR_primers/          # Primer sequences for validation
├── README.md              # ← You are here

---

👨‍🔬 Acknowledgements

This project was developed under collaboration between:

• Australian Institute for Bioengineering and Nanotechnology (AIBN)
• University of Queensland (UQ)

---

🔗 Contact

For questions or collaborations, please contact:
📧 [[email protected]]