Skip to content

Cluster Analysis Specifications

Process Details

  • Name: Cluster Analysis
  • Process UUID: 86744cb86c004fe59eebb77b71dc9bd6
  • Process Group: Misc.

Overview

This process performs RNA secondary structure prediction and visualization for clustered RNA sequences using reactivity data. It integrates external constraints such as DMS, SHAPE, or PAIR-MaP reactivity data with RNA sequences to predict secondary structures for each cluster, and generates visualization plots of the clustering results.

This process is implemented in Bash, which invokes Python scripts (foldClusters.py and plotClusters.py) for structure prediction and cluster visualization respectively.

Key Functionality

  • Secondary Structure Prediction: Performs RNA secondary structure folding for each cluster using RNAstructure algorithms with integrated reactivity constraints
  • Reactivity Data Integration: Incorporates experimental reactivity data (DMS, SHAPE, PAIR-MaP) as constraints during folding predictions
  • Cluster Visualization: Generates plots showing reactivity patterns and clustering results with customizable parameters
  • Multiple Input Format Support: Handles both BM files and reactivity text files as input data sources

Input/Output Specification

Inputs

Required Inputs

  • Directory
    • Description: Input directory containing RNA reactivity data files and associated sequence information
    • Format: Directory

Outputs

  • Directory

    • Description: Output directory containing predicted secondary structures and analysis results
    • Format: Directory

  • Image

    • Description: Visualization plot showing cluster analysis results and reactivity patterns
    • Format: PNG

Parameters & Settings

These parameters can be adjusted in the Foundry UI when running this process.

  • Data Type

    • Description: Choose the data type from which a plot will be drawn. norm: Normalize (Default), raw: Raw data
    • Available options: norm (default), raw

  • Input Type

    • Description: Please choose the input type. Could be BM (.bm) or REACT files (reactivities.txt)
    • Available options: bm (default), react

  • Cluster Folding Parameters

    • Description: Define the rest of the parameters for foldClusters.py through here. --bp [string]: Prefix for bp files. I.e. if your pairmap files are test-[]-pairmap.bp, you should pass '--bp test', --prob: Compute pairing probability (partition), --pk: Fold using ShapeKnots (default is to use Fold), --nog: Mask Gs (set to no-data), --nou: Mask Us (set to no-data)
    • Default value: (empty)

  • Cluster Plotting Parameters

    • Description: Define the rest of the parameters for plotClusters.py through here. --bm1 [BM1]: Path of first bm file, --bm2 [BM2]: Path of second bm file (optional), --react1 [REACT1]: Plot -reactivities.txt file (instead of BM file), --react2 [REACT2]: Path of second reactivities file, --ptype [PTYPE]: Which data to plot from reactivities file. Can be norm/raw, --out [OUT]: Write figure to file. If --out flag not used, plot using interactive GUI, --align: Align bm1 and bm2 based on reactivity (rather than population), --models [MODELS]: Only plot the specified models. E.g. --models 0,2 will plot models 0 and 2 for a 3 component model. Only works when visualizing single bm or react.
    • Default value: (empty)

References & Resources

  • Tool Documentation: Contact the team for details on foldClusters.py and plotClusters.py