# ggplotAgent: a self-debugging multi-modal agent for robust and reproducible scientific visualization

**Authors:** Zelin Wang, Yuanyuan Yin, Jien Wang, Haiyan Yan, Xuan Xie, Yiqing Zheng

PMC · DOI: 10.1093/bioadv/vbaf332 · Bioinformatics Advances · 2026-01-02

## TL;DR

ggplotAgent is an AI tool that automatically creates high-quality scientific visualizations from natural language prompts, reducing the need for coding expertise.

## Contribution

The novel self-debugging multi-modal agent improves code execution and aesthetic accuracy for scientific visualizations.

## Key findings

- ggplotAgent achieved a 100% code executability rate compared to 85% for DeepSeek-V3.
- It produced a 'Publication-Ready' score of 1.9 versus 0.7 for the baseline model.
- The agent enhanced plots beyond user prompts, achieving a positive Insight Score of +0.3.

## Abstract

Creating publication-quality visualizations is essential for bioinformatics but remains a bottleneck for researchers with limited coding expertise. While Large Language Models (LLMs) are proficient at generating code, they often fail in practice due to library dependencies, dataset mismatches, or syntax errors. These issues require manual intervention, slowing data interpretation.

We present ggplotAgent, a novel multi-modal, self-debugging artificial intelligence agent that automates publication-ready ggplot2 visualizations. It features a dual-layered framework that resolves code execution errors and uses a vision-enabled agent to verify aesthetic correctness. In benchmarks against the DeepSeek-V3 model, ggplotAgent achieved a 100% code executability rate(versus 85%) and a “Publication-Ready” score of 1.9 (versus 0.7). Surprisingly, it showcased the ability to act as an expert collaborator by intelligently enhancing plots beyond the user’s literal prompt, achieving a positive Insight Score of +0.3 over than the baseline (−0.05). These results demonstrate its ability to reliably produce accurate, high-quality visualizations directly from natural language.

ggplotAgent is freely accessible as a public web application at https://ggplotagent.databio1.com/ and an offline Streamlit app. The source code is available on GitHub at https://github.com/charlin90/ggplotAgent. This software is distributed under the MIT License.

## Full text

_Full body text omitted from this summary view._ Fetch the complete paper as Markdown: https://tomesphere.com/paper/PMC12802885/full.md

## Figures

1 figure with captions in the complete paper: https://tomesphere.com/paper/PMC12802885/full.md

## References

9 references — full list in the complete paper: https://tomesphere.com/paper/PMC12802885/full.md

---
Source: https://tomesphere.com/paper/PMC12802885