inst/QUICK_REFERENCE.md

In FracFixR: Compositional Statistical Framework for RNA Fractionation Analysis

FracFixR Quick Reference Guide

Installation

# From CRAN
install.packages("FracFixR")

# From GitHub
devtools::install_github("Arnaroo/FracFixR")

Basic Workflow

1. Load Package and Data

library(FracFixR)

# Your data
counts <- as.matrix(read.csv("counts.csv", row.names = 1))
annotation <- read.csv("annotation.csv")

2. Run FracFixR

results <- FracFixR(MatrixCounts = counts, 
                    Annotation = annotation)

3. Visualize Fractions

PlotFractions(results)

4. Differential Testing

diff_results <- DiffPropTest(results,
                            Conditions = c("Control", "Treatment"),
                            Types = "Heavy_Polysome",
                            Test = "GLM")

5. Volcano Plot

PlotComparison(diff_results,
               Conditions = c("Control", "Treatment"),
               Types = "Heavy_Polysome")

Key Functions

| Function | Purpose | Key Parameters | |----------|---------|----------------| | FracFixR() | Main analysis | MatrixCounts, Annotation | | DiffPropTest() | Differential testing | NormObject, Conditions, Types, Test | | PlotFractions() | Visualize proportions | FracFixed | | PlotComparison() | Volcano plot | DiffPropResult, Conditions, Types |

Annotation Requirements

Required columns: - Sample: Must match column names in count matrix - Condition: Experimental conditions - Type: Fraction types (must include "Total") - Replicate: Replicate identifiers

Example:

Sample    Condition  Type    Replicate
Sample1   Control    Total   Rep1
Sample2   Control    Light   Rep1
Sample3   Control    Heavy   Rep1

Statistical Tests

• GLM (default): Most powerful, binomial GLM
• Logit: Faster, logit transformation
• Wald: Beta-binomial for overdispersed data

Common Analyses

Single Fraction

diff_heavy <- DiffPropTest(results,
                          Conditions = c("A", "B"),
                          Types = "Heavy")

Combined Fractions

diff_combined <- DiffPropTest(results,
                             Conditions = c("A", "B"),
                             Types = c("Light", "Heavy"))

Filter Results

# Significant genes
sig_genes <- diff_results[diff_results$padj < 0.01, ]

# Top changed genes
top_genes <- diff_results[order(abs(diff_results$mean_diff), 
                                decreasing = TRUE), ][1:50, ]

Output Components

FracFixR() returns: - $OriginalData: Filtered input counts - $Annotation: Sample annotation - $Propestimates: Proportion estimates - $NewData: Corrected counts - $Coefficients: Regression coefficients - $Fractions: Fraction proportions - $plots: Diagnostic plots

DiffPropTest() returns: - transcript: Gene/transcript ID - mean_success_cond1/2: Mean proportions - mean_diff: Difference in proportions - log2FC: Log2 fold change - pval: Raw p-value - padj: FDR-adjusted p-value

Tips

1. Check Total samples: Ensure each condition-replicate has a "Total"
2. Minimum replicates: Use ≥2 replicates per condition
3. Parallel processing: Automatic, uses available cores
4. Memory: For >50k genes, pre-filter low counts
5. Interpretation: Positive mean_diff = higher in condition 2

Example Datasets

# Polysome profiling
data(example_counts)
data(example_annotation)

# Alternative annotations
data(polysome_annotation)    # Monosome/Polysome
data(subcellular_annotation)  # Nuclear/Cytoplasmic

Troubleshooting

See full troubleshooting guide in package documentation or:

vignette("FracFixR-intro")
help(FracFixR)

Citation

citation("FracFixR")

Cleynen et al. (2024) FracFixR: A compositional statistical framework for absolute proportion estimation between fractions in RNA sequencing data. Bioinformatics.

FracFixR documentation built on May 11, 2026, 9:09 a.m.