Thymela

Thymela is a web application designed to analyze and search biomedical articles, with a particular focus on proteomics research data. The project provides a comprehensive suite of tools for scraping, analyzing, and searching research articles and their associated metadata. It was developed in collaboration with Meta-Flux @ DogpatchLabs!

NOTE - This is a prototype developed over the course of a week and has not been fully tested. As a result, proper formatting and 'best practices' are yet to be implemented.

Features

• Web Scraping: Automated scraping of research articles from multiple sources including:

  • Metabolomics Workbench (MWB)
  • PRIDE database
  • MetaboLights

• Metadata Analysis:

  • Extracts and processes metadata from research articles
  • Handles various data formats and structures
  • Post-processing capabilities for different data sources

• Search Functionality:

  • search capabilities using vector indexing (via Pinecone! <3)

Project Structure

src/
├── analyse_articles.py     # Core article analysis functionality
├── dbwrap/                 # Database wrapper and operations
├── parsing/                # Article parsing and API integration
├── prompting/              # Prompt management for analysis
├── searching/              # Search functionality implementation
├── templates/              # HTML templates
├── webapp/                 # Web application components
└── webscraping/            # Web scraping modules