miRNA Target Consensus Demo

This repository is a synthetic portfolio demo showing how predictions from multiple miRNA target-prediction tools can be normalized, integrated, summarized, and converted into consensus-based target tables.

The project uses only mock data. It does not contain unpublished research data, real thesis results, real biological candidates, real sample names, or experimental validation.

Overview

miRNA target-prediction workflows often produce outputs from multiple tools with different column names, scoring systems, and formats. This demo shows a simple reproducible workflow to:

• read mock prediction outputs from multiple tools;
• normalize predictions into a common long-format table;
• group predictions by mirna_id and target_id;
• calculate the number of tools supporting each interaction;
• classify interactions into simple consensus levels;
• join mock gene annotations;
• export consensus tables, gene lists, and summary tables.

Repository Structure

mirna-target-consensus-demo/
├── README.md
├── requirements.txt
├── LICENSE
├── .gitignore
├── data/
│   └── mock/
│       ├── mock_mirnas.fasta
│       ├── mock_miranda_predictions.tsv
│       ├── mock_rnahybrid_predictions.tsv
│       ├── mock_psrnatarget_predictions.tsv
│       ├── mock_ptarpmir_predictions.tsv
│       └── mock_gene_annotation.tsv
├── src/
│   ├── normalize_predictions.py
│   ├── build_consensus.py
│   ├── generate_gene_lists.py
│   ├── summarize_support.py
│   └── run_demo.py
├── results/
│   └── expected/
└── docs/
    └── workflow.md

Input Data

All input files are synthetic and stored in:

data/mock/

The mock inputs simulate outputs from four miRNA target-prediction tools:

• miRanda
• RNAhybrid
• psRNATarget
• PTarPmiR

Every prediction file contains at least:

• mirna_id
• target_id

Tool-specific columns are preserved as an evidence field during normalization.

Installation

Create and activate a Python environment:

python3 -m venv .venv
source .venv/bin/activate

Install dependencies:

pip install -r requirements.txt

Run the Demo

From the repository root:

python src/run_demo.py

Expected Outputs

The workflow generates:

results/expected/normalized/all_predictions_long.tsv
results/expected/consensus/mirna_target_consensus.tsv
results/expected/gene_lists/
results/expected/summary/tool_support_summary.tsv
results/expected/summary/consensus_level_summary.tsv

Consensus Logic

For each mirna_id and target_id pair:

• tool_support_n = number of prediction tools supporting the interaction;
• tools_supporting = semicolon-separated list of tools;
• consensus_level:
  • high = supported by 3 or more tools;
  • medium = supported by 2 tools;
  • single_tool = supported by 1 tool.

Workflow Scripts

• src/normalize_predictions.py: converts heterogeneous mock prediction files into one standardized long-format table.
• src/build_consensus.py: groups predictions by miRNA-target pair and calculates consensus support.
• src/generate_gene_lists.py: exports one target-gene list per mock miRNA.
• src/summarize_support.py: generates summary tables by tool and consensus level.
• src/run_demo.py: runs the complete workflow in order.

Limitations

This repository is a portfolio demonstration. It does not provide biological validation.

Important limitations:

• all data are synthetic/mock;
• scores are illustrative and not biologically meaningful;
• consensus support does not prove that a target is real;
• real miRNA target analysis requires careful tool selection, parameter control, reference versioning, biological interpretation, and experimental validation;
• this demo does not perform real target prediction or wet-lab validation.

Author

Enrique A. Caban Centeno
PhD researcher and bioinformatics analyst

Focus areas: RNA/miRNA analysis, sequence analysis, biological annotation, qPCR data processing, reproducible Python/R workflows, and omics-style data analysis.

License

This project is released under the MIT License. See the LICENSE file for details.

Portfolio Relevance

This repository was created as a public portfolio example for RNA/miRNA bioinformatics workflows. It demonstrates how heterogeneous prediction outputs can be standardized, integrated, summarized, and converted into consensus-based result tables using a reproducible Python workflow.

The project is suitable for demonstrating skills in Python, pandas, bioinformatics data processing, multi-tool result integration, target prioritization logic, and scientific workflow documentation.

A concise portfolio-oriented summary is available in:

docs/portfolio_summary.md