Differential Protein Abundance Results

Published

Jun 2026

After result-table quality control, the next step is to filter proteins based on differential abundance criteria.

This chapter moves from understanding differential abundance columns to producing interpretable protein lists.

The goal is to identify proteins that pass both:

statistical evidence
        +
biological effect size

In this guide, the default filtering criteria are:

adjusted_p_value <= 0.05
absolute log2fc >= 1

These thresholds are practical defaults for the running example. Real projects may require different thresholds depending on the study design, sample size, platform, and biological question.

Starting Point

This chapter uses the running example result table:

data/example/example-proteomics-results.csv

Before running this chapter, the table should have passed the basic QC checks from Chapter 04.

The Chapter 04 command was:

Rscript scripts/R/04-qc-result-table.R \
  data/example/example-proteomics-results.csv \
  results

The QC output should include:

results/result-table-qc-summary.tsv

Why Filtering Matters

A proteomics result table may contain many proteins.

Some proteins have strong statistical evidence. Some have large fold changes. Some have both.

Filtering helps identify proteins that are most useful for downstream biological interpretation.

The filtering logic in this chapter is:

adjusted_p_value <= threshold
        ↓
statistically supported

absolute log2fc >= threshold
        ↓
biologically meaningful change

both criteria satisfied
        ↓
differentially abundant protein

Direction of Change

The direction of change depends on the comparison design.

For the running example:

treated_vs_control

means:

numerator condition   = treated
denominator condition = control

Therefore:

positive log2fc → higher abundance in treated
negative log2fc → lower abundance in treated

The script classifies proteins as:

upregulated
downregulated
not_significant

In this results-first proteomics guide, upregulated means higher abundance in the numerator condition of the comparison. It does not necessarily imply transcriptional regulation.

Expected Input

The expected input is:

data/example/example-proteomics-results.csv

The table should contain:

protein_id
gene_symbol
protein_name
log2fc
p_value
adjusted_p_value
comparison

The script can also work with real project input:

data/input/proteomics-results.csv

Output Files

The filtering script creates:

results/
├── differential-proteins.tsv
├── significant-proteins.tsv
├── upregulated-proteins.tsv
├── downregulated-proteins.tsv
└── differential-summary.tsv

Filtering Script

The executable script is saved as:

scripts/R/05-filter-differential-proteins.R

The guide shows how to run the script. The full executable code is maintained in scripts/R/.

Running the Script on the Example Input

From the project root, run:

Rscript scripts/R/05-filter-differential-proteins.R \
  data/example/example-proteomics-results.csv \
  results \
  1 \
  0.05

The arguments are:

1st argument → input table
2nd argument → output directory
3rd argument → absolute log2 fold-change threshold
4th argument → adjusted p-value threshold

In this example:

absolute log2fc threshold = 1
adjusted p-value threshold = 0.05

Running the Script on a Real Project Input

For a real project table, run:

Rscript scripts/R/05-filter-differential-proteins.R \
  data/input/proteomics-results.csv \
  results \
  1 \
  0.05

Only the input file changes.

Output 1: Differential Proteins

The full differential table is:

results/differential-proteins.tsv

This file contains all proteins from the input table, plus filtering and interpretation columns.

Added columns include:

log2fc_numeric
adjusted_p_value_numeric
abs_log2fc
passes_padj
passes_log2fc
is_significant
regulation

This file is useful because it preserves the full result table while making the filtering decision explicit.

Output 2: Significant Proteins

The significant protein table is:

results/significant-proteins.tsv

This file contains only proteins that pass both filtering criteria:

adjusted_p_value <= threshold
absolute log2fc >= threshold

This is the main table for downstream interpretation, annotation, enrichment, and visualization.

Output 3: Upregulated Proteins

The upregulated protein table is:

results/upregulated-proteins.tsv

This file contains significant proteins with positive log2 fold change.

For the running example:

positive log2fc → higher abundance in treated

Output 4: Downregulated Proteins

The downregulated protein table is:

results/downregulated-proteins.tsv

This file contains significant proteins with negative log2 fold change.

For the running example:

negative log2fc → lower abundance in treated

or equivalently:

higher abundance in control

Output 5: Differential Summary

The differential summary file is:

results/differential-summary.tsv

It summarizes:

total proteins
significant proteins
upregulated proteins
downregulated proteins
not significant proteins
thresholds used

This file provides a quick checkpoint for the analysis.

Expected Results from the Example Table

Using the running example and the default thresholds:

absolute log2fc >= 1
adjusted_p_value <= 0.05

the example should produce a small set of significant proteins.

This gives a simple confirmation that the workflow is working before larger real datasets are introduced.

Why Keep All Proteins?

Even after filtering, it is useful to keep the full differential table.

The complete table is needed for:

reviewing proteins near the threshold
changing thresholds later
ranking all proteins
creating volcano plots
preparing background sets for enrichment analysis
documenting the full decision process

For that reason, this chapter saves both:

all differential results
filtered significant proteins

Thresholds Are Analysis Decisions

The thresholds used here are not fixed biological truths.

They are analysis decisions.

A stricter analysis might use:

adjusted_p_value <= 0.01
absolute log2fc >= 1.5

A more exploratory analysis might use:

adjusted_p_value <= 0.10
absolute log2fc >= 0.58

The important point is that the thresholds should be stated clearly and recorded in the output.

Looking Ahead

The next chapter focuses on ranking and filtering differentially expressed or differentially abundant proteins.

Filtering identifies proteins that pass thresholds.

Ranking prioritizes proteins for interpretation.

Differential protein abundance filtering
        ↓
Significant proteins
        ↓
Upregulated and downregulated proteins
        ↓
Ranked protein lists
        ↓
Biological interpretation