GeneGuessr: A New Daily Biology Puzzle Game

A new daily web puzzle named GeneGuessr has been launched for molecular biology enthusiasts. Modeled after popular games like Geoguessr and Wordle, the game challenges players to identify a specific human protein based on a 3D model and a series of similarity clues. The objective is to triangulate the correct gene name within a limited number of guesses.

The developer, a professional with a background in wet lab molecular biology, created the application over a two-month period. While the primary target audience is fellow biologists, the creator is also testing the game's accessibility for non-experts, specifically those who might utilize browser-based Large Language Models (LLMs) to assist in solving the daily puzzle.

The core mechanic of GeneGuessr involves analyzing a 3D representation of a human protein. Players must use provided similarity clues to deduce the correct gene name. The game is designed to be a daily ritual, offering a new challenge each day similar to the format established by Wordle.

The creator's intent is to make the game engaging primarily for other biologists. However, there is a secondary interest in how the game performs for those outside the scientific field. The developer specifically invites non-biologists to attempt solving the puzzles, particularly if they employ browser-based AI tools to aid their guesses. This approach aims to determine if the game serves as an educational tool for a broader audience.

The game is accessible via a direct URL and requires no user registration or login, lowering the barrier to entry. The developer has noted that the mobile version may contain untested bugs, as the primary development focus was on desktop functionality.

The creation of GeneGuessr highlights the growing capability of AI in software development. The creator, who describes their coding experience as limited to basic Python data analysis and figure generation, built the entire web application using Claude over the course of two months.

To manage the development workflow, the creator utilized specific AI-assisted tools. Linear MCP was instrumental in organizing the project, allowing the AI to place individual issues on a shared Kanban board. Additionally, Playwright MCP was used for testing the application on a live site.

When facing complex bugs that Claude could not resolve in a single attempt, the creator used Linear to consolidate issue information. This information was then fed into ChatGPT Codex, which, despite providing confusing explanations, successfully resolved the bugs after processing for approximately thirty minutes.

The project serves as a direct response to a common question observed in online communities: "Now that we have coding AI, why isn't there a deluge of awesome AI-generated apps made by non-coders?" GeneGuessr is presented as a concrete example of what a web application looks like when built by a non-coder leveraging modern AI tools.

By sharing the development process, the creator demonstrates that complex, functional web apps are now within reach for individuals without traditional software engineering backgrounds. The reliance on Claude for coding and specialized MCPs for project management illustrates a new paradigm in application development.

The creator has expressed openness to writing more about the development process if there is sufficient interest from the community. This suggests that the project is not just a finished product, but also a case study in the evolving relationship between human domain experts and AI coding assistants.