Functional Profiling

Published

Jun 2026

ID: MICROB-009
Type: System Component
Audience: Students, researchers, analysts, and practitioners
Theme: Exploring microbial functional potential

Introduction

Functional profiling is the stage where microbiome analysis moves from microbial identity toward microbial potential.

Taxonomic profiling asks:

Who is there?

Functional profiling asks:

What might the microbial community be able to do?

In shotgun metagenomics, functional profiling may summarize genes, gene families, enzymes, pathways, modules, or other functional units. In marker-gene studies, functional potential may sometimes be inferred, but such inference should be interpreted cautiously.

In the Microbiome Analysis System, functional profiling is treated as a complementary layer of evidence. It can strengthen biological interpretation, but it should not be overinterpreted as direct evidence of gene expression, protein activity, or metabolic flux.

Why Functional Profiling Matters

Functional profiling helps connect microbial communities to biological processes.

It can support questions such as:

Which microbial pathways are present?
Are functional profiles different across samples or groups?
Do taxonomic patterns correspond to functional patterns?
Are specific gene families or pathways enriched in one condition?
Do observed functions support a biological hypothesis?

Functional profiling is especially useful when the biological question concerns metabolism, host-microbe interactions, nutrient cycling, resistance genes, virulence potential, or environmental function.

Position in the Microbiome Analysis System

Functional profiling occurs after feature generation and can support downstream statistical analysis and biological interpretation.

Show code

flowchart LR
  A[Feature Generation] --> B[Functional Profiling]
  B --> C[Differential Analysis]
  B --> D[Biological Interpretation]
  B --> E[Reproducible Reporting]

flowchart LR
  A[Feature Generation] --> B[Functional Profiling]
  B --> C[Differential Analysis]
  B --> D[Biological Interpretation]
  B --> E[Reproducible Reporting]

Functional profiling should be interpreted with the sequencing strategy in mind.

Functional Potential vs Functional Activity

Functional profiling usually measures or infers functional potential.

Functional potential means that genes or pathways are detected or predicted in the community.

Functional activity requires additional evidence, such as:

metatranscriptomics
metaproteomics
metabolomics
enzyme activity assays
experimental validation

A pathway detected in a metagenomic profile does not necessarily mean that the pathway is active under the sampled condition.

Functional Units

Functional profiling may summarize several types of features.

Common functional units include:

gene families
enzyme commission numbers
KEGG orthologs
MetaCyc pathways
Gene Ontology terms
antimicrobial resistance genes
virulence factors
carbohydrate-active enzymes
metabolic modules

The correct unit depends on the biological question and the profiling method.

Functional Profiling From Marker-Gene Data

Marker-gene data, such as 16S rRNA sequencing, primarily measure taxonomic composition.

Some methods infer functional potential from taxonomy, but these predictions depend heavily on reference genomes and assumptions about functional similarity among related organisms.

For marker-gene data, inferred functional profiles should be reported as predictions, not direct measurements.

Functional Profiling From Shotgun Metagenomics

Shotgun metagenomics can directly profile functional potential because sequencing reads are sampled across genomes rather than only a marker gene.

Shotgun functional profiling may involve:

quality control
host-read removal
read classification
gene family profiling
pathway reconstruction
normalization
statistical comparison
biological interpretation

Shotgun workflows can provide richer functional information, but they also require careful handling of databases, parameters, and computational resources.

Example Functional Profiling Scripts

The following scripts provide a lightweight MAS-side functional profiling example.

These scripts do not perform real shotgun metagenomic functional profiling. They create and summarize a small toy pathway abundance table so the MAS workflow can continue into differential analysis and interpretation.

The workflow uses two scripts:

scripts/bash/09a-create-example-functional-profile.sh
scripts/R/09b-plot-functional-profile.R

09a: Create the Example Functional Profile

Save this script as:

scripts/bash/09a-create-example-functional-profile.sh

#!/bin/bash

###############################################################################
# Microbiome Analysis System
# 09a-create-example-functional-profile.sh
#
# Purpose:
#   Create a small example functional profile for workflow testing.
#
# Important:
#   This script creates toy pathway abundances and mock annotations.
#   These are not real functional profiling results and should not be used for
#   biological interpretation.
#
# Usage:
#   bash scripts/bash/09a-create-example-functional-profile.sh
###############################################################################

set -e

FUNCTION_DIR="data/function"
REPORT_DIR="data/reports"

mkdir -p "${FUNCTION_DIR}"
mkdir -p "${REPORT_DIR}"

FUNCTION_TABLE="${FUNCTION_DIR}/pathway-abundance.tsv"
FUNCTION_METADATA="${FUNCTION_DIR}/pathway-metadata.tsv"
REPORT_FILE="${REPORT_DIR}/functional-profile-report.tsv"

echo "Creating MAS example functional profile..."

cat > "${FUNCTION_TABLE}" <<'EOF'
pathway_id  SRR17868090 SRR17868091 SRR17868092
PWY_001 45  38  22
PWY_002 12  18  35
PWY_003 5   7   20
PWY_004 25  21  19
PWY_005 3   5   12
EOF

cat > "${FUNCTION_METADATA}" <<'EOF'
pathway_id  pathway_name    category    notes
PWY_001 Carbohydrate fermentation   Metabolism  toy pathway
PWY_002 Short-chain fatty acid production   Metabolism  toy pathway
PWY_003 Bile acid transformation    Host-microbe interaction    toy pathway
PWY_004 Amino acid biosynthesis Metabolism  toy pathway
PWY_005 Oxidative stress response   Stress response toy pathway
EOF

pathway_count=$(tail -n +2 "${FUNCTION_TABLE}" | wc -l | tr -d ' ')
sample_count=$(head -n 1 "${FUNCTION_TABLE}" | awk -F '\t' '{print NF-1}')

printf "metric\tvalue\n" > "${REPORT_FILE}"
printf "pathway_count\t%s\n" "${pathway_count}" >> "${REPORT_FILE}"
printf "sample_count\t%s\n" "${sample_count}" >> "${REPORT_FILE}"
printf "profile_type\ttoy pathway abundance profile\n" >> "${REPORT_FILE}"
printf "functional_profile_status\tREADY_FOR_PLOTTING\n" >> "${REPORT_FILE}"

echo "Example functional profile created."
echo
echo "Created:"
echo "  ${FUNCTION_TABLE}"
echo "  ${FUNCTION_METADATA}"
echo "  ${REPORT_FILE}"
echo
echo "Next:"
echo "  Rscript scripts/R/09b-plot-functional-profile.R"

Run it from the MAS project root:

bash scripts/bash/09a-create-example-functional-profile.sh

This creates:

data/function/pathway-abundance.tsv
data/function/pathway-metadata.tsv
data/reports/functional-profile-report.tsv

09b: Plot the Functional Profile

Save this script as:

scripts/R/09b-plot-functional-profile.R

###############################################################################
# Microbiome Analysis System
# 09b-plot-functional-profile.R
#
# Purpose:
#   Plot a simple pathway-level functional profile.
#
# Usage:
#   Rscript scripts/R/09b-plot-functional-profile.R
###############################################################################

library(readr)
library(dplyr)
library(tidyr)
library(ggplot2)

function_dir <- "data/function"
figure_dir <- "figures"
table_dir <- "tables"

dir.create(figure_dir, recursive = TRUE, showWarnings = FALSE)
dir.create(table_dir, recursive = TRUE, showWarnings = FALSE)

pathway_file <- file.path(function_dir, "pathway-abundance.tsv")
metadata_file <- file.path(function_dir, "pathway-metadata.tsv")

if (!file.exists(pathway_file)) {
  stop("Missing pathway abundance file: ", pathway_file,
       "\nRun: bash scripts/bash/09a-create-example-functional-profile.sh")
}

if (!file.exists(metadata_file)) {
  stop("Missing pathway metadata file: ", metadata_file,
       "\nRun: bash scripts/bash/09a-create-example-functional-profile.sh")
}

pathways <- read_tsv(pathway_file, show_col_types = FALSE)
pathway_metadata <- read_tsv(metadata_file, show_col_types = FALSE)

pathway_long <- pathways %>%
  pivot_longer(
    cols = -pathway_id,
    names_to = "sample_id",
    values_to = "abundance"
  ) %>%
  left_join(pathway_metadata, by = "pathway_id") %>%
  group_by(sample_id) %>%
  mutate(
    relative_abundance = abundance / sum(abundance),
    relative_abundance_percent = relative_abundance * 100
  ) %>%
  ungroup()

write_tsv(
  pathway_long,
  file.path(table_dir, "pathway-relative-abundance-for-plot.tsv")
)

p <- ggplot(
  pathway_long,
  aes(
    x = sample_id,
    y = relative_abundance_percent,
    fill = pathway_name
  )
) +
  geom_col() +
  labs(
    title = "Example Functional Profile",
    subtitle = "Toy MAS pathway abundance data for workflow testing",
    x = "Sample",
    y = "Relative abundance (%)",
    fill = "Pathway"
  ) +
  theme_minimal(base_size = 12) +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

ggsave(
  filename = file.path(figure_dir, "pathway-relative-abundance-profile.png"),
  plot = p,
  width = 8,
  height = 5,
  dpi = 300
)

message("Created:")
message("  ", file.path(table_dir, "pathway-relative-abundance-for-plot.tsv"))
message("  ", file.path(figure_dir, "pathway-relative-abundance-profile.png"))

Run it from the MAS project root:

Rscript scripts/R/09b-plot-functional-profile.R

This creates:

tables/pathway-relative-abundance-for-plot.tsv
figures/pathway-relative-abundance-profile.png

Running the Complete Functional Profiling Example

To create and plot the example functional profile, run:

bash scripts/bash/09a-create-example-functional-profile.sh
Rscript scripts/R/09b-plot-functional-profile.R
cat data/reports/functional-profile-report.tsv

The example functional table is intentionally small and artificial. It is designed to demonstrate workflow structure, not biological function.

Example Pathway Abundance Table

The generated pathway table looks like this:

pathway_id  SRR17868090 SRR17868091 SRR17868092
PWY_001 45  38  22
PWY_002 12  18  35
PWY_003 5   7   20
PWY_004 25  21  19
PWY_005 3   5   12

This table can support downstream demonstration of functional summaries and differential analysis structure.

Interpreting Functional Profiles

Functional profiles should be interpreted in relation to:

sequencing strategy
functional profiling method
reference database
normalization method
sample metadata
taxonomic context
statistical testing
biological plausibility

Functional profiling can support biological interpretation, but it should not be treated as direct evidence of activity unless supported by additional data.

Common Functional Profiling Issues

Common issues include:

treating inferred functions as measured functions
ignoring database limitations
comparing profiles generated by different tools
overinterpreting pathway names
ignoring normalization and compositionality
failing to distinguish gene presence from gene expression
reporting broad functions without biological context
ignoring uncertainty in annotation

These issues should be documented in the final report.

MAS Functional Profiling Outputs

At the end of this stage, MAS should have:

functional abundance table
functional metadata table
functional profile report
plot-ready functional abundance table
functional profile figure
notes on functional profiling method and limitations

Show code

flowchart LR
  A[Feature or Metagenomic Data] --> B[Functional Profile]
  B --> C[Pathway Abundance Table]
  C --> D[Functional Plot]
  D --> E[Biological Interpretation]

flowchart LR
  A[Feature or Metagenomic Data] --> B[Functional Profile]
  B --> C[Pathway Abundance Table]
  C --> D[Functional Plot]
  D --> E[Biological Interpretation]

Key Takeaways

Functional profiling adds a biological process layer to microbiome analysis.

A strong functional profiling stage ensures that:

functional units are clearly defined
pathway or gene family abundance tables are documented
database and method limitations are reported
functional potential is not confused with activity
results are interpreted alongside taxonomy and metadata

Functional profiling can strengthen biological interpretation, but only when its assumptions and limitations are clearly documented.

What Comes Next

The next chapter examines Differential Analysis, where microbiome features, taxa, or functional profiles are compared across groups or conditions.

# Functional Profiling :::cdi-message - **ID:** MICROB-009 - **Type:** System Component - **Audience:** Students, researchers, analysts, and practitioners - **Theme:** Exploring microbial functional potential ::: ## Introduction Functional profiling is the stage where microbiome analysis moves from microbial identity toward microbial potential. Taxonomic profiling asks: ```text Who is there? ``` Functional profiling asks: ```text What might the microbial community be able to do? ``` In shotgun metagenomics, functional profiling may summarize genes, gene families, enzymes, pathways, modules, or other functional units. In marker-gene studies, functional potential may sometimes be inferred, but such inference should be interpreted cautiously. In the Microbiome Analysis System, functional profiling is treated as a complementary layer of evidence. It can strengthen biological interpretation, but it should not be overinterpreted as direct evidence of gene expression, protein activity, or metabolic flux. ## Why Functional Profiling Matters Functional profiling helps connect microbial communities to biological processes. It can support questions such as: - Which microbial pathways are present? - Are functional profiles different across samples or groups? - Do taxonomic patterns correspond to functional patterns? - Are specific gene families or pathways enriched in one condition? - Do observed functions support a biological hypothesis? Functional profiling is especially useful when the biological question concerns metabolism, host-microbe interactions, nutrient cycling, resistance genes, virulence potential, or environmental function. ## Position in the Microbiome Analysis System Functional profiling occurs after feature generation and can support downstream statistical analysis and biological interpretation. ```{mermaid} flowchart LR A[Feature Generation] --> B[Functional Profiling] B --> C[Differential Analysis] B --> D[Biological Interpretation] B --> E[Reproducible Reporting] ``` Functional profiling should be interpreted with the sequencing strategy in mind. ## Functional Potential vs Functional Activity Functional profiling usually measures or infers functional potential. Functional potential means that genes or pathways are detected or predicted in the community. Functional activity requires additional evidence, such as: - metatranscriptomics - metaproteomics - metabolomics - enzyme activity assays - experimental validation A pathway detected in a metagenomic profile does not necessarily mean that the pathway is active under the sampled condition. ## Functional Units Functional profiling may summarize several types of features. Common functional units include: - gene families - enzyme commission numbers - KEGG orthologs - MetaCyc pathways - Gene Ontology terms - antimicrobial resistance genes - virulence factors - carbohydrate-active enzymes - metabolic modules The correct unit depends on the biological question and the profiling method. ## Functional Profiling From Marker-Gene Data Marker-gene data, such as 16S rRNA sequencing, primarily measure taxonomic composition. Some methods infer functional potential from taxonomy, but these predictions depend heavily on reference genomes and assumptions about functional similarity among related organisms. For marker-gene data, inferred functional profiles should be reported as predictions, not direct measurements. ## Functional Profiling From Shotgun Metagenomics Shotgun metagenomics can directly profile functional potential because sequencing reads are sampled across genomes rather than only a marker gene. Shotgun functional profiling may involve: - quality control - host-read removal - read classification - gene family profiling - pathway reconstruction - normalization - statistical comparison - biological interpretation Shotgun workflows can provide richer functional information, but they also require careful handling of databases, parameters, and computational resources. ## Example Functional Profiling Scripts The following scripts provide a lightweight MAS-side functional profiling example. These scripts do not perform real shotgun metagenomic functional profiling. They create and summarize a small toy pathway abundance table so the MAS workflow can continue into differential analysis and interpretation. The workflow uses two scripts: ```text scripts/bash/09a-create-example-functional-profile.sh scripts/R/09b-plot-functional-profile.R ``` ## 09a: Create the Example Functional Profile Save this script as: ```bash scripts/bash/09a-create-example-functional-profile.sh ``` ```bash #!/bin/bash ############################################################################### # Microbiome Analysis System # 09a-create-example-functional-profile.sh # # Purpose: # Create a small example functional profile for workflow testing. # # Important: # This script creates toy pathway abundances and mock annotations. # These are not real functional profiling results and should not be used for # biological interpretation. # # Usage: # bash scripts/bash/09a-create-example-functional-profile.sh ############################################################################### set -e FUNCTION_DIR="data/function" REPORT_DIR="data/reports" mkdir -p "${FUNCTION_DIR}" mkdir -p "${REPORT_DIR}" FUNCTION_TABLE="${FUNCTION_DIR}/pathway-abundance.tsv" FUNCTION_METADATA="${FUNCTION_DIR}/pathway-metadata.tsv" REPORT_FILE="${REPORT_DIR}/functional-profile-report.tsv" echo "Creating MAS example functional profile..." cat > "${FUNCTION_TABLE}" <<'EOF' pathway_id SRR17868090 SRR17868091 SRR17868092 PWY_001 45 38 22 PWY_002 12 18 35 PWY_003 5 7 20 PWY_004 25 21 19 PWY_005 3 5 12 EOF cat > "${FUNCTION_METADATA}" <<'EOF' pathway_id pathway_name category notes PWY_001 Carbohydrate fermentation Metabolism toy pathway PWY_002 Short-chain fatty acid production Metabolism toy pathway PWY_003 Bile acid transformation Host-microbe interaction toy pathway PWY_004 Amino acid biosynthesis Metabolism toy pathway PWY_005 Oxidative stress response Stress response toy pathway EOF pathway_count=$(tail -n +2 "${FUNCTION_TABLE}" | wc -l | tr -d ' ') sample_count=$(head -n 1 "${FUNCTION_TABLE}" | awk -F '\t' '{print NF-1}') printf "metric\tvalue\n" > "${REPORT_FILE}" printf "pathway_count\t%s\n" "${pathway_count}" >> "${REPORT_FILE}" printf "sample_count\t%s\n" "${sample_count}" >> "${REPORT_FILE}" printf "profile_type\ttoy pathway abundance profile\n" >> "${REPORT_FILE}" printf "functional_profile_status\tREADY_FOR_PLOTTING\n" >> "${REPORT_FILE}" echo "Example functional profile created." echo echo "Created:" echo " ${FUNCTION_TABLE}" echo " ${FUNCTION_METADATA}" echo " ${REPORT_FILE}" echo echo "Next:" echo " Rscript scripts/R/09b-plot-functional-profile.R" ``` Run it from the MAS project root: ```bash bash scripts/bash/09a-create-example-functional-profile.sh ``` This creates: ```text data/function/pathway-abundance.tsv data/function/pathway-metadata.tsv data/reports/functional-profile-report.tsv ``` ## 09b: Plot the Functional Profile Save this script as: ```text scripts/R/09b-plot-functional-profile.R ``` ```r ############################################################################### # Microbiome Analysis System # 09b-plot-functional-profile.R # # Purpose: # Plot a simple pathway-level functional profile. # # Usage: # Rscript scripts/R/09b-plot-functional-profile.R ############################################################################### library(readr) library(dplyr) library(tidyr) library(ggplot2) function_dir <- "data/function" figure_dir <- "figures" table_dir <- "tables" dir.create(figure_dir, recursive = TRUE, showWarnings = FALSE) dir.create(table_dir, recursive = TRUE, showWarnings = FALSE) pathway_file <- file.path(function_dir, "pathway-abundance.tsv") metadata_file <- file.path(function_dir, "pathway-metadata.tsv") if (!file.exists(pathway_file)) { stop("Missing pathway abundance file: ", pathway_file, "\nRun: bash scripts/bash/09a-create-example-functional-profile.sh") } if (!file.exists(metadata_file)) { stop("Missing pathway metadata file: ", metadata_file, "\nRun: bash scripts/bash/09a-create-example-functional-profile.sh") } pathways <- read_tsv(pathway_file, show_col_types = FALSE) pathway_metadata <- read_tsv(metadata_file, show_col_types = FALSE) pathway_long <- pathways %>% pivot_longer( cols = -pathway_id, names_to = "sample_id", values_to = "abundance" ) %>% left_join(pathway_metadata, by = "pathway_id") %>% group_by(sample_id) %>% mutate( relative_abundance = abundance / sum(abundance), relative_abundance_percent = relative_abundance * 100 ) %>% ungroup() write_tsv( pathway_long, file.path(table_dir, "pathway-relative-abundance-for-plot.tsv") ) p <- ggplot( pathway_long, aes( x = sample_id, y = relative_abundance_percent, fill = pathway_name ) ) + geom_col() + labs( title = "Example Functional Profile", subtitle = "Toy MAS pathway abundance data for workflow testing", x = "Sample", y = "Relative abundance (%)", fill = "Pathway" ) + theme_minimal(base_size = 12) + theme(axis.text.x = element_text(angle = 45, hjust = 1)) ggsave( filename = file.path(figure_dir, "pathway-relative-abundance-profile.png"), plot = p, width = 8, height = 5, dpi = 300 ) message("Created:") message(" ", file.path(table_dir, "pathway-relative-abundance-for-plot.tsv")) message(" ", file.path(figure_dir, "pathway-relative-abundance-profile.png")) ``` Run it from the MAS project root: ```bash Rscript scripts/R/09b-plot-functional-profile.R ``` This creates: ```text tables/pathway-relative-abundance-for-plot.tsv figures/pathway-relative-abundance-profile.png ``` ## Running the Complete Functional Profiling Example To create and plot the example functional profile, run: ```bash bash scripts/bash/09a-create-example-functional-profile.sh Rscript scripts/R/09b-plot-functional-profile.R cat data/reports/functional-profile-report.tsv ``` The example functional table is intentionally small and artificial. It is designed to demonstrate workflow structure, not biological function. ## Example Pathway Abundance Table The generated pathway table looks like this: ```text pathway_id SRR17868090 SRR17868091 SRR17868092 PWY_001 45 38 22 PWY_002 12 18 35 PWY_003 5 7 20 PWY_004 25 21 19 PWY_005 3 5 12 ``` This table can support downstream demonstration of functional summaries and differential analysis structure. ## Interpreting Functional Profiles Functional profiles should be interpreted in relation to: - sequencing strategy - functional profiling method - reference database - normalization method - sample metadata - taxonomic context - statistical testing - biological plausibility Functional profiling can support biological interpretation, but it should not be treated as direct evidence of activity unless supported by additional data. ## Common Functional Profiling Issues Common issues include: - treating inferred functions as measured functions - ignoring database limitations - comparing profiles generated by different tools - overinterpreting pathway names - ignoring normalization and compositionality - failing to distinguish gene presence from gene expression - reporting broad functions without biological context - ignoring uncertainty in annotation These issues should be documented in the final report. ## MAS Functional Profiling Outputs At the end of this stage, MAS should have: - functional abundance table - functional metadata table - functional profile report - plot-ready functional abundance table - functional profile figure - notes on functional profiling method and limitations ```{mermaid} flowchart LR A[Feature or Metagenomic Data] --> B[Functional Profile] B --> C[Pathway Abundance Table] C --> D[Functional Plot] D --> E[Biological Interpretation] ``` ## Key Takeaways Functional profiling adds a biological process layer to microbiome analysis. A strong functional profiling stage ensures that: - functional units are clearly defined - pathway or gene family abundance tables are documented - database and method limitations are reported - functional potential is not confused with activity - results are interpreted alongside taxonomy and metadata Functional profiling can strengthen biological interpretation, but only when its assumptions and limitations are clearly documented. ## What Comes Next The next chapter examines **Differential Analysis**, where microbiome features, taxa, or functional profiles are compared across groups or conditions.