Karel Berka - Personal Homepage

This repository contains the source code for the personal and academic homepage of Prof. Karel Berka, Ph.D.

🔗 Live Webpage: https://karelberka.github.io/

---

Directory Structure

To keep the repository clean and manageable, files are organized as follows:

karelberka.github.io/
│
├── index.html                   # Main page (About, Outreach, Selected Publications, Contact)
├── publication.html             # Dynamic publication details detail page
├── .gitignore                   # Files excluded from Git tracking
├── README.md                    # Project documentation
│
├── css/
│   └── style.css                # Styling (Light/Dark themes, responsive layout)
│
├── js/
│   ├── script.js                # Core interactive logic (Canvas background, filters, contact form)
│   └── data.js                  # Publication catalog (JSON array)
│
├── scripts/
│   └── generate_missing_thumbnails.py   # Python utility script to generate publication art via Gemini
│
└── img/
    ├── profile/                 # Profile pictures, favicon caricatures, and signature files
    ├── projects/                # Logo and screenshots for research platforms (MOLEonline, ChannelsDB)
    └── publications/            # Individual publication thumbnails (named as pub_[Year]_[Journal]_[Author].png)

---

Technical Details

• Frontend: Pure HTML5, CSS3, and Vanilla JavaScript (zero external JS dependencies).
• Interactivity: Includes a responsive molecular-dynamics simulation canvas background that interacts with mouse hover coordinates.
• Language Support: Dynamic English and Czech translation toggle via localStorage.
• Publications Detail: Subpage loads details dynamically from js/data.js based on query parameters (e.g., publication.html?id=10.1093/nar/gkag511).

---

Generation of Publication Art

A python utility script is available at scripts/generate_missing_thumbnails.py. It reads a metadata file (enriched_works.json) and queries the Gemini API (imagen-4.0-generate-001) to generate beautiful, scientific-oriented AI illustrations for newly added publications.

Running the Generator

1. Install the Google GenAI SDK:

   pip install google-genai

2. Configure your API key:

   export GEMINI_API_KEY="your-api-key-here"

3. Run the script:

   python scripts/generate_missing_thumbnails.py

It will automatically search for publication thumbnails in img/publications/ that are small placeholders (~68 bytes transparent files) and replace them with high-quality generated illustrations.