← Về thư mục
📄 / / profiles / r-and-d / raw_evidence / OpenAI4S / skills / bioskills / bio-ribo-seq-orf-detection / references / usage-guide.md

ORF Detection - Usage Guide

Overview

Detect actively translated open reading frames from Ribo-seq data using 3-nucleotide periodicity (not coverage) as the evidence of translation: canonical CDS, uORFs, internal ORFs, dORFs, and novel ORFs. Classify them by the 2022 community standard, quantify ORF-level translation, and validate calls before trusting them.

Prerequisites

pip install RiboCode

# Isoform-aware quantification and the general toolkit (distinct packages)
BiocManager::install(c("ORFquant", "ORFik", "DESeq2"))

Quick Start

Tell your AI agent what you want to do: - "Find translated ORFs in my Ribo-seq data with RiboCode" - "Detect uORFs including near-cognate (CUG/GUG) starts" - "Quantify ORF-level translation isoform-aware with ORFquant" - "Validate my called ORFs (in-frame fraction, FLOSS)"

Example Prompts

ORF Detection

"Run RiboCode on my transcriptome BAM and report ORFs with adjusted p < 0.05"

"Include CUG and GTG near-cognate start codons in the uORF search"

"How many novel ORFs and uORFs are detected?"

Classification and Nomenclature

"Classify my ORFs by the Mudge 2022 categories (uORF, uoORF, intORF, dORF, doORF, lncRNA-ORF)"

"Find translated ORFs on long non-coding RNAs"

Quantification and Validation

"Quantify ORF-level translation isoform-aware with ORFquant"

"Compute FLOSS and in-frame fraction to validate my called ORFs"

"Compare ORF-level translation between conditions with DESeq2"

What the Agent Will Do

  1. Prepare transcript annotation and run metaplots to select periodic read lengths
  2. Call ORFs with RiboCode (optionally with near-cognate starts)
  3. Classify ORFs by the standard positional categories
  4. Quantify isoform-aware with ORFquant or count P-sites per ORF
  5. Validate with in-frame fraction, FLOSS, and conservation/peptide evidence

Tips