Skip to content

RSEM Count Specifications

Process Details

  • Name: RSEM Count
  • Process UUID: 20bdc34a-ed05-11e8-9ce4-85221b21564c
  • Process Group: Misc.

Overview

This process merges RSEM (RNA-Seq by Expectation Maximization) output files into single table files for downstream analysis. It processes multiple RSEM result files from different samples and consolidates them into unified expression matrices, facilitating comparative analysis across samples.

This process is implemented in Bash, which invokes a Perl script for file processing and data aggregation.

Key Functionality

  • Expression Matrix Generation: Creates consolidated expression matrices from individual RSEM output files
  • Multi-Level Processing: Processes both gene-level and isoform-level expression data
  • Multiple Metrics Support: Handles expected counts and TPM (Transcripts Per Million) values
  • Strand-Specific Analysis: Supports forward, reverse, and unstranded expression quantification
  • Automated File Discovery: Automatically identifies and processes RSEM output files in the working directory

Input/Output Specification

Inputs

Required Inputs

  • rsemOut
    • Description: RSEM output files containing gene and isoform expression quantification results
    • Format: rsemOut

Outputs

  • outputFile
    • Description: Consolidated expression matrix tables containing merged expression data across all samples
    • Format: TSV

Parameters & Settings

No user-configurable parameters are available for this process in the Foundry UI.

References & Resources

  • Tool Documentation: Contact the team for details on the Perl processing script
  • Related Papers: Li, B., & Dewey, C. N. (2011). RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome. BMC bioinformatics, 12(1), 323.