Robin: an advanced tool for comparative loop caller result analysis leveraging large language models

TL;DR

Robin is a web-based platform that uses large language models to analyze and compare chromosome loop detection results, helping researchers validate and visualize genomic data.

Contribution

Robin introduces a novel platform that integrates large language models with genomic analysis and visualization tools for real-time loop caller result comparison.

Findings

Robin provides a comprehensive set of metrics for analyzing and comparing loop caller results.

The platform integrates HiGlass for interactive visualization of Hi-C matrices and loop annotations.

LLM capabilities allow users to generate customized plots using natural language instructions.

Abstract

There has been significant interest in genomics research, leading to the development of numerous new methods. One notable area of progress is in chromosome loop detection algorithms (also known as loop callers). However, despite these advancements, there is no available platform to analyze, compare, or benchmark current tools’ results on the go. Developing such a platform is crucial to accelerating research and ensuring the reliability and effectiveness of new methods in this field. Hence, in this work, we propose Robin, an advanced, ready-to-go platform for comparative loop caller result analysis that leverages large language models (LLMs). Robin is a web server designed to analyze loop caller results, offering a comprehensive range of analysis metrics, such as recovery and overlap. It is tightly integrated with HiGlass for interactive, multi-resolution visualization of Hi-C matrices and loop annotations, allowing users to visually inspect and validate loop structures in a genomic context. Additionally, Robin incorporates LLM capabilities that enable users to generate customized plots and figures simply by providing natural language instructions. Overall, Robin is a robust and comprehensive loop caller result analysis and visualization tool. It is publicly accessible at http://hicrobin.online, with comprehensive documentation available at http://documentation.hicrobin.online/.

Figures