BioinfoMCP is an open-source platform released in April 2026 by the Florensia Widjaja research team at the Chinese University of Hong Kong, Shenzhen (CUHK Shenzhen). It automatically converts existing command-line interface (CLI)-based bioinformatics analysis tools into a Model Context Protocol (MCP) server, allowing the latest AI agents to directly control them and verifying their integrity. This system acts as an intelligent translator, connecting complex genomics analysis programs with AI models. It consists of two core layers: a Converter that automatically codes the MCP server based on tool manuals, and a Benchmark Suite that verifies the communication integrity and parameter mapping of the converted server. Existing bioinformatics CLI tools, widely used in genomic analysis, have numerous command-line parameters and complex dependency structures. When controlled by large language models (LLMs) that communicate only in natural language, they often cause syntax errors or apply incorrect options, leading to hallucinations. Similar to how an autonomous driving system drives a car's steering and accelerator as physical control signals, BioinfoMCP builds a standardized neural network-machine interface that allows AI agents communicating in natural language to directly and accurately control legacy CLI bioinformatics analysis tools without errors. This ensures that the agent can mechanically execute bioinformatics commands without errors through a predefined, formalized schema, even without knowing complex shell syntax. Biotechnology researchers can use this platform to analyze large-scale next-generation sequencing (NGS) data, driving the entire pipeline with intuitive natural language queries without the cumbersome bash scripting process. For example, a complex multi-step analysis process, from FastQC quality control to Bowtie2 sequence alignment and MACS3 transcription factor binding site detection, can be performed by the agent autonomously with a single request. In particular, by providing only the tool's terminal help (--help) output or PDF-format technical documentation, the AI can package code that can be directly operated within minutes, dramatically reducing the time and cost of adding new genomic analysis pipelines to a lab's unique infrastructure and automating experiments.

BioinfoMCP is an open-source platform released in April 2026 by the Florensia Widjaja research team at the Chinese University of Hong Kong, Shenzhen (CUHK Shenzhen). It automatically converts existing command-line interface (CLI)-based bioinformatics analysis tools into a Model Context Protocol (MCP) server, allowing the latest AI agents to directly control them and verify their integrity. This system enables complex genomic analysis programs and artificial intelligence.