Kaidoku AI Agent

What it does

• Answers questions about bioinformatics concepts and file formats
• Knows all Kaidoku tools and pipelines — with step-by-step guidance
• Suggests tools and pipelines that match your goal (e.g. FastQC → RNA-Seq → differential expression)
• Can run select analyses in chat (alignment, motifs, phylogeny, GC content, NCBI search, and more)
• Can discuss uploaded sequence or result snippets you paste in chat (when enabled)

Model & approach

Responses are generated by a large language model configured for scientific tone. Outputs are assistive—not a substitute for peer review, clinical validation, or domain expert sign-off where required.

What it can see

Kaidoku AI Agent sees what you put in the chat (text and optional file excerpts you attach in the product). It may use session context such as tool names or pipeline labels shown in the UI to give relevant guidance.

What it cannot see

• Patient identifiers or protected health information unless you paste them—do not upload PHI
• Other users' data or runs
• Your full account or billing details unless you explicitly describe them in chat

Data retention

Chat content may be processed by the model provider according to our subprocessors and your plan. Do not submit secrets or regulated health data. For institutional requirements, contact us about deployment options.

How to use it well

• State organism, read type (DNA/RNA), and goal (e.g. "variant calling from exome FASTQ")
• Paste small excerpts of errors or QC stats rather than entire files
• Ask for step-by-step reasoning and verify against documentation

Example queries

• "My FASTQ mean quality drops at read ends—should I trim before STAR?"
• "Explain N50 vs L50 for my assembly summary."
• "What inputs does the metagenomics pipeline need for Kraken2?"

Limitations

Kaidoku AI Agent can hallucinate or be outdated versus the latest tool versions. Always check parameter docs and cite primary literature for publications. Not for medical diagnosis or treatment decisions.