Gene-Genie

A web API and client for aggregating genomic information from multiple data sources.

Overview

Imagine if every appliance you purchased came with its own unique power cord requiring custom wiring to fit your outlets. Instead, appliances are designed to plug into standardized outlets — no special knowledge required. This is what an API accomplishes for data.

Genomics data presents a particular challenge: extreme variability in both the data itself and the formats used to represent it. Gene-Genie addresses this by consolidating information from multiple API sources into a unified interface, allowing you to query and cross-reference genomic databases without navigating their underlying complexity.

Data Sources:

• Entrez — Access to PubMed, PMC, Gene, and Nuccore databases
• String-db.org — Protein annotations and network interactions
• UniProt — Resource ID tags and cross-references
• GenomeWeb — Genomics news and articles

Features

• Search genes by name and retrieve functional annotations
• Cross-reference identifiers across multiple databases
• Fetch related articles from PubMed
• Display protein-protein interaction networks
• Tag cloud visualization of database cross-references

Architecture

The project consists of two components:

• API Layer — Flask-based REST API that wraps external genomics data sources
• Web Client — Flask application with a search interface styled using Bulma

Note: A migration to FastAPI is planned. See Roadmap for details.

Installation

Prerequisites

• Python 3.8+
• pip

Setup

# Clone the repository
git clone https://github.com/<your-username>/gene_genie.git
cd gene_genie

# Create and activate virtual environment
python3 -m venv venv
source ./venv/bin/activate

# Install dependencies
pip install -r requirements.txt

Usage

Running Locally

cd gene_genie
source ./venv/bin/activate

export FLASK_APP=gene_genie.py
export FLASK_ENV=development

flask run -h localhost -p 8000

The web client will be available at http://localhost:8000/.

Web Client

The web client provides a search interface for querying genomic data:

Search Page

• Header — Navigation with link to home
• Search — Gene name entry form
• Results — Display area for query results

Results Page

• Annotations — Functional annotations from String-db.org
• Tags — Cross-reference IDs to external databases
• Related Articles — Publications from PubMed
• Interaction Graph — Protein network visualization (when available)

API Reference

HTTP Method	Endpoint	Description
POST	/api/v1.0/stringdb/annotation/<term>	Retrieve annotation from String-db.org
POST	/api/v1.0/stringdb/annotations/<term>	Retrieve multiple annotations from String-db.org
POST	/api/v1.0/uniprot/<term>	Retrieve resource ID tags from UniProt
GET	/api/v1.0/pubmed/summary/<pmid>	Retrieve summary article from PubMed
GET	/api/v1.0/genomeweb/article/<gwid>	Retrieve single news item from GenomeWeb
GET	/api/v1.0/genomeweb/articles/	Retrieve multiple news items from GenomeWeb

Example Requests

# String-db annotations
curl -i http://localhost:8000/api/v1.0/stringdb/annotation/BRCA1
curl -i http://localhost:8000/api/v1.0/stringdb/annotations/BRCA1

# UniProt cross-references
curl -i http://localhost:8000/api/v1.0/uniprot/BRCA1

# PubMed article summary
curl -i http://localhost:8000/api/v1.0/pubmed/summary/31411802

# GenomeWeb news
curl -i http://localhost:8000/api/v1.0/genomeweb/article/1
curl -i http://localhost:8000/api/v1.0/genomeweb/articles/

Example Responses

UniProt cross-references (/api/v1.0/uniprot/BRCA1):

{
  "EMBL": "AF355273",
  "GO": "GO:0016567",
  "ID": "Q90Z51_CHICK",
  "KEGG": "gga:373983",
  "KO": "K10605",
  "PROSITE": "PS50089",
  "Pfam": "PF00097",
  "Reactome": "R-GGA-351433",
  "RefSeq": "NP_989500.1",
  "SMART": "SM00184"
}

PubMed summary (/api/v1.0/pubmed/summary/<pmid>):

{
  "PMID": "31411802",
  "Title": "BRCA mutation frequency and clinical features of ovarian...",
  "Author": "Bu H",
  "Journal": "J Obstet Gynaecol Res",
  "Abstract": "Subjects with germline BRCA1/2 mutations (gBRCAm) have...",
  "Citation": "Bu H - J Obstet Gynaecol Res",
  "Url": "http://www.ncbi.nlm.nih.gov/pubmed/31411802"
}

GenomeWeb article (/api/v1.0/genomeweb/article/<gwid>):

{
  "feature_items": [
    {
      "date": "Aug 20, 2019",
      "link": "USPSTF Formalizes Slightly Expanded BRCA Screening Recommendations",
      "id": 1
    }
  ]
}

User Stories

These scenarios illustrate how Gene-Genie serves different users in a genomics workflow.

Translate Gene IDs Across Databases

A developer building a bioinformatics pipeline needs to map an Entrez GeneID to corresponding identifiers in UniProt, KEGG, and other databases. Rather than querying each resource separately, she calls the UniProt endpoint with a gene name and receives a unified set of cross-references.

curl -i http://localhost:8000/api/v1.0/uniprot/BRCA1

Retrieve Related Literature

A graduate student researching a specific gene wants to find recently published articles. Using a PubMed ID obtained from previous results, he retrieves article metadata including title, authors, abstract, and a direct link to the publication.

curl -i http://localhost:8000/api/v1.0/pubmed/summary/31411802

Explore Interaction Networks

The same student recalls that BRCA1 interacts with other genes as part of a network. The String-db endpoint returns functional annotations and can visualize protein-protein interaction networks, helping clarify the gene's role in broader biological pathways.

curl -i http://localhost:8000/api/v1.0/stringdb/annotations/BRCA1

Access Genomics News

A non-technical user wants to learn more about genomics in the news. The GenomeWeb endpoint provides access to curated articles, making current developments accessible without requiring domain expertise.

curl -i http://localhost:8000/api/v1.0/genomeweb/articles/

Deployment

Cloud Deployment (AWS EC2)

# Install dependencies on EC2 instance
sudo apt-get install python3-venv

# Clone and set up the project
git clone https://github.com/<your-username>/gene_genie.git
cd gene_genie
python3 -m venv venv
source ./venv/bin/activate
pip install -r requirements.txt

# Run the application
./run.sh

The application will be accessible at http://<your-ec2-public-ip>:8000/.

Tip: For production deployments, consider using Gunicorn with Nginx as a reverse proxy.

Development

Running API Modules Directly

Individual API modules can be run standalone for testing and development:

source ./venv/bin/activate

# Entrez API
./app/api/entrez_api.py

# String-db API
./app/api/stringdb_api.py

# UniProt API
./app/api/uniprot_api.py

# PubMed API
./app/api/pubmed_api.py

GenomeWeb API (interactive):

from app.api import genome_web_api

gw_api = genome_web_api.GenomeWebAPI()
items = gw_api.fetch()
print(items)

Roadmap

Framework Migration:

• Migrate from Flask to FastAPI
• Add OpenAPI/Swagger documentation (automatic with FastAPI)

Async & Performance:

• Implement async support for concurrent API requests
• Add Celery task queue with Redis backend for request scheduling
• Implement Ajax for browser-based async calls
• Add response caching

Infrastructure:

• Containerize with Docker
• Add Docker Compose for local development

Testing & Quality:

• Add unit tests (expand Flask-Testing coverage)
• Add integration tests for external API endpoints
• Set up CI/CD pipeline

License

MIT License