Package 'TraianProt'

Description

Generates a bar plot from an enrichResult object, faceted by cluster.

Usage

barplot_func(terms, number, conditions, ...)

Arguments

Value

A ggplot object.

Examples

# This example requires a valid enrichResult object.
# We will create a minimal synthetic one.
if (requireNamespace("DOSE") && requireNamespace("methods")) {
    dummy_data <- data.frame(
        Cluster = c("Up-regulated", "Down-regulated"),
        Category = "BP", ID = c("GO:0006915", "GO:0007049"),
        Description = c("apoptosis", "cell cycle"),
        p.value = c(0.01, 0.02), adj.P.Val = c(0.01, 0.02),
        query_size = c(50, 40), Count = c(5, 8), term_size = c(100, 120),
        effective_domain_size = 10000, geneID = c("P04637", "P10415"),
        GeneRatio = c("5/50", "8/40"), BgRatio = c("100/10000", "120/10000"),
        Conditions = c("Up-regulated", "Down-regulated")
    )
    dummy_enrich_result <- methods::new("enrichResult", result = dummy_data)
    terms_list <- list(NULL, dummy_enrich_result)
    barplot_func(terms_list, number = 1, conditions = "My Plot")
}

Description

Generates a ggplot2 boxplot of LOG2 intensities, grouped by sample and colored by condition.

Usage

boxplot_function(df, metadata, selected_conditions)

Arguments

Value

A ggplot object.

Examples

# Create synthetic data
log_df <- data.frame(
    LOG2.C1 = rnorm(20, 10),
    LOG2.C2 = rnorm(20, 10.1),
    LOG2.T1 = rnorm(20, 12),
    LOG2.T2 = rnorm(20, 12.1)
)

meta <- data.frame(
    group = c("Control", "Control", "Treatment", "Treatment"),
    log2_col = c("LOG2.C1", "LOG2.C2", "LOG2.T1", "LOG2.T2")
)

# Generate the boxplot
boxplot_function(log_df, meta, selected_conditions = c("Control", "Treatment"))

Description

This function identifies technical replicates based on a biological replicate metadata column and collapses them by calculating the mean intensity. It updates both the expression data and the associated metadata.

Usage

collapse_technical_replicates(df, metadata)

Arguments

Details

The function assumes that technical replicates share the same sample_name ID in the metadata. It cleans the column names by removing trailing numeric indices (e.g., "Sample, 1" becomes "Sample") using regex. If only one technical replicate exists for a biological group, it simply renames the column.

Value

A list with two elements:

• data: A data.frame with collapsed biological replicates as columns.

• metadata: A data.frame with one row per biological replicate, updated to match the new column names.

Examples

# Example metadata
meta <- data.frame(
    log2_col = c("S1, 1", "S1, 2", "S2, 1"),
    sample_name = c("Bio1", "Bio1", "Bio2"),
    intensity_sample_name = c("S1, 1", "S1, 2", "S2, 1"),
    stringsAsFactors = FALSE
)
# Example data
dat <- data.frame(
    `S1, 1` = c(1, 2), `S1, 2` = c(1.2, 2.2), `S2, 1` = c(5, 6),
    check.names = FALSE
)
result <- collapse_technical_replicates(dat, meta)

Description

Calculates a Pearson correlation matrix from a data frame and visualizes it as a tile-based heatmap using ggplot2, overlaying the correlation coefficients.

Usage

corrplot_function(
  df,
  metadata,
  display = "circle",
  tl.col = "black",
  addCoef.col = "black"
)

Arguments

Details

Generate a ggplot2 Correlation Matrix Heatmap

The function first calculates the correlation matrix using cor() with use = "complete.obs", which handles missing values by pairwise deletion. It then melts the matrix into a "long" format suitable for ggplot2. The plot uses scale_fill_gradient2 to create a divergent color scale (Red -> White -> Blue) centered at 0. coord_fixed() ensures square tiles.

Value

A ggplot object representing the correlation heatmap.

Examples

df_sample <- data.frame(
    LOG2.C1 = rnorm(20, mean = 10),
    LOG2.C2 = rnorm(20, mean = 10),
    LOG2.T1 = rnorm(20, mean = 12),
    LOG2.T2 = rnorm(20, mean = 12)
)

meta <- data.frame(
    group = c("Control", "Control", "Treatment", "Treatment"),
    log2_col = c("LOG2.C1", "LOG2.C2", "LOG2.T1", "LOG2.T2")
)

# 2. Generate the plot
p <- corrplot_function(df_sample, meta, addCoef.col = "blue")

# 3. Display the plot (Important: in examples we just call the object)
p

Description

Creates a boxplot with jittered points for a single protein, comparing two conditions.

Usage

Diferential_boxplot(
  df,
  metadata,
  protein = NULL,
  LOG2.names,
  selected_conditions
)

Arguments

Value

This function is called for its side effect of printing a ggplot. It invisibly returns the ggplot object p.

Examples

# Create synthetic data
df_box <- data.frame(
    Protein = "Prot1",
    LOG2.C1 = rnorm(1, 8), LOG2.C2 = rnorm(1, 8),
    LOG2.T1 = rnorm(1, 9), LOG2.T2 = rnorm(1, 9)
)
metadata_box <- data.frame(
    log2_col = c("LOG2.C1", "LOG2.C2", "LOG2.T1", "LOG2.T2"),
    group = c("Control", "Control", "Treatment", "Treatment")
)
log_names <- c("LOG2.C1", "LOG2.C2", "LOG2.T1", "LOG2.T2")
selected_cond <- c("Treatment", "Control")

# Generate the plot
if (requireNamespace("DOSE")) {
    Diferential_boxplot(df_box, metadata_box, "Prot1", log_names, selected_cond)
}

Description

Generates a dot plot from a compareClusterResult object, faceted by condition.

Usage

dotplot_func(terms, ...)

Arguments

Value

A ggplot object.

Examples

if (requireNamespace("clusterProfiler", quietly = TRUE)) {
    dummy_data <- data.frame(
        Cluster = "Up-regulated",
        Category = "BP",
        ID = "GO:0006915",
        Description = "apoptosis",
        p.value = 0.01,
        p.adjust = 0.01,
        Count = 5,
        GeneRatio = "5/50",
        BgRatio = "100/10000",
        Conditions = "Up-regulated"
    )

    dummy_cc_result <- methods::new("compareClusterResult",
        compareClusterResult = dummy_data
    )
    dotplot_func(list(dummy_cc_result))
}

Description

A helper function to filter a protein expression data frame (e.g., from MaxQuant) based on a minimum number of valid (non-NA, non-zero) values across samples.

Usage

filter_valids(
  df,
  metadata,
  unique_proteins,
  min_prop = NULL,
  at_least_one = FALSE,
  labeltype = 1
)

Arguments

Value

A list containing the filtered data frame ("filtered") and the LOG2-transformed data frame ("LOG2").

Examples

raw_df <- data.frame(
    Protein = c("P1", "P2", "P3", "P4"),
    LOG2_A1 = c(20, 21, NA, 22),
    LOG2_A2 = c(20, NA, NA, 22),
    LOG2_B1 = c(25, 24, 23, NA),
    LOG2_B2 = c(25, 24, 23, 26)
)
meta_df <- data.frame(
    intensity_sample_name = c("A1", "A2", "B1", "B2"),
    group = c("Control", "Control", "Treatment", "Treatment"),
    log2_col = c("LOG2_A1", "LOG2_A2", "LOG2_B1", "LOG2_B2")
)
unique_list <- list(
    data.frame(Protein = character()),
    data.frame(Protein = character())
)

filtered_data <- filter_valids(
    df = raw_df,
    metadata = meta_df,
    unique_proteins = unique_list,
    min_prop = 0.5,
    at_least_one = FALSE,
    labeltype = 1
)

head(filtered_data)

Description

This function is a helper wrapper that extracts the raw gprofiler2 results (the third element) from the output of the Goterms_finder function to generate a Manhattan plot of enriched terms.

Usage

gostplot_func(terms, ...)

Arguments

Value

A ggplot or plotly object representing the Manhattan plot of functional enrichment.

Examples

gostplot_func(res)

Description

Performs Gene Ontology (GO) and pathway enrichment analysis for up- and down-regulated proteins using gprofiler2.

Usage

Goterms_finder(
  df,
  raw,
  target,
  numeric_ns,
  mthreshold,
  filter_na,
  organismo,
  custombg,
  platform,
  ...
)

Arguments

Value

A list named go_structures containing: 1. A compareClusterResult object (for plotting). 2. An enrichResult object (for plotting). 3. The raw gost result object (for gostplot).

Examples

limma_df <- data.frame(
    Protein = c("P04637", "P00533", "P01116"),
    Protein_description = c("TP53", "EGFR", "KRAS"),
    expression = c("Up-regulated", "Up-regulated", "Up-regulated"),
    p.value = c(0.001, 0.002, 0.003)
)

# raw_df is needed for the function structure, even if custombg=FALSE
raw_df <- data.frame(
    Protein = c("P04637", "P00533", "P01116", "P62258"),
    Protein_description = c("TP53", "EGFR", "KRAS", "ACTB")
)

# 2. Run the function with checks
# This example requires an internet connection

if (requireNamespace("gprofiler2", quietly = TRUE)) {
    try(
        {
            go_res <- Goterms_finder(
                df = limma_df,
                raw = raw_df,
                target = "ENSG",
                numeric_ns = "",
                mthreshold = 0.05,
                filter_na = TRUE,
                organismo = "hsapiens",
                custombg = FALSE,
                platform = 1
            )

            # 3. Safe printing (checks if result exists before printing)
            # We check if go_res exists, is a list, and has the expected slot
            if (!is.null(go_res) &&
                is.list(go_res) &&
                length(go_res) >= 1 &&
                inherits(go_res[[1]], "compareClusterResult")) {
                print(head(as.data.frame(go_res[[1]])))
            }
        },
        silent = TRUE
    )
}

Description

A simple wrapper to plot a histogram for a specific column in a data frame.

Usage

histogram(df, colname, color, title)

Arguments

Value

A histogram plot.

Examples

my_data <- data.frame(p.value = rnorm(100))
histogram(my_data, "p.value", "lightblue", "P-Value Distribution")

Description

Identifies unique and common proteins between two data frames.

Usage

identify_proteins(raw, metadata, platform, selected_conditions)

Arguments

Value

A list containing three data frames: unique_df1 (proteins unique to df1), unique_df2 (proteins unique to df2), and common_df (proteins common to both).

Examples

metadata_example <- data.frame(
    intensity_sample_name = c("Log_C1", "Log_C2", "Log_T1", "Log_T2"),
    group = c("Control", "Control", "Treatment", "Treatment")
)

raw_example <- data.frame(
    Protein = c("Prot_A", "Prot_B", "Prot_C", "Prot_D"),
    Log_C1 = c(25.1, 24.2, 0, 22.0),
    Log_C2 = c(25.0, 24.5, 0, 22.2),
    Log_T1 = c(0, 26.1, 23.5, 21.9),
    Log_T2 = c(0, 26.0, 23.8, 22.1)
)
plot_result <- identify_proteins(
    raw = raw_example,
    metadata = metadata_example,
    platform = 1,
    selected_conditions = c("Control", "Treatment")
)
print(plot_result)

Description

Calculates graph theory measures for up- and down-regulated protein networks.

Usage

igraph_analysis(interactions, taxonid, score)

Arguments

Value

A list of two data frames, Upregulated and Downregulated, containing graph measures for each network.

Examples

if (requireNamespace("STRINGdb") && requireNamespace("igraph")) {
    # Create synthetic data
    up_mapped <- data.frame(
        Protein = c("TP53", "EGFR"),
        STRING_id = c("9606.ENSP00000269305", "9606.ENSP00000275493")
    )
    down_mapped <- data.frame(
        Protein = c("CALCA", "ACTB"),
        STRING_id = c("9606.ENSP00000302220", "9606.ENSP00000482455")
    )
    interactions_list <- list(up_mapped, down_mapped)

    # Run analysis
    analysis_results <- igraph_analysis(interactions_list, 9606, 400)
    print(analysis_results$Upregulated)
}

Description

Replaces non-finite (NA, Inf) values in specified LOG2 columns with random values drawn from a normal distribution. This distribution is shifted to a lower mean and narrower standard deviation relative to the observed valid data.

Usage

impute_data(df, LOG2.names, width = 0.3, downshift = 1.8)

Arguments

Details

This method is a common "down-shift" or "Perseus-style" imputation for label-free proteomics, simulating the low-abundance signals of proteins near the detection limit. It assumes missingness is not at random (MNAR).

Value

The data frame df with imputed LOG2 columns and new 'impute.*' logical columns indicating which values were imputed.

Examples

# Create synthetic LOG2 data with non-finite values (Inf)
log_df <- data.frame(
    Protein = c("ProtA", "ProtB", "ProtC", "ProtD"),
    LOG2.S1 = c(10, 12, Inf, 8),
    LOG2.S2 = c(11, 13, Inf, Inf),
    # Add a KEEP column (e.g., from a filter)
    # Here, we keep all proteins for calculation
    KEEP = c(TRUE, TRUE, TRUE, TRUE)
)

# Specify columns to impute
log_cols <- c("LOG2.S1", "LOG2.S2")

# Run imputation
set.seed(123) # for reproducible example
imputed_df <- impute_data(log_df, log_cols)

print("Original Data:")
print(log_df)
print("Imputed Data:")
print(imputed_df)

Description

Uses K-Nearest Neighbors (k-NN) to impute missing (NA) values in a log-expression matrix.

Usage

impute_KNN_data(df, LOG2.names, ...)

Arguments

Value

A numeric matrix with NA values imputed.

Examples

log_matrix_na <- matrix(
    c(
        10.0, 10.2, NA, 10.5, 10.1, 9.9, # Sample 1 (with NAs)
        12.0, NA, 11.9, 12.1, 11.8, NA
    ), # Sample 2 (with NAs)
    nrow = 6, ncol = 2,
    dimnames = list(
        c("P1", "P2", "P3", "P4", "P5", "P6"),
        c("Sample1", "Sample2")
    )
)

LOG2.names <- c("Sample1", "Sample2")

imputed_matrix <- impute_KNN_data(as.data.frame(log_matrix_na), LOG2.names)

print(imputed_matrix)

Description

Maps down-regulated proteins to the STRING database and plots the network.

Usage

interactions_down(df, taxonid, score)

Arguments

Value

This function is called for its side effect of generating a plot. It does not return an object.

Examples

if (requireNamespace("STRINGdb")) {
    # Create synthetic data
    limma_df <- data.frame(
        Protein = c("P10415", "P62258"), # CALCA, ACTB
        expression = c("Down-regulated", "Down-regulated")
    )
    # Run function
    interactions_down(limma_df, taxonid = 9606, score = 400)
}

Description

Maps up-regulated proteins to the STRING database, plots the network, and returns a list of mapped proteins.

Usage

interactions_up(df, taxonid, score)

Arguments

Value

A list containing two data frames: 1. up_mapped: Mapped up-regulated proteins. 2. down_mapped: Mapped down-regulated proteins.

Examples

if (requireNamespace("STRINGdb")) {
    # Create synthetic data
    limma_df <- data.frame(
        Protein = c("P04637", "P00533"), # TP53, EGFR
        expression = c("Up-regulated", "Up-regulated")
    )
    # Run function
    interactions_list <- interactions_up(limma_df, taxonid = 9606, score = 400)
}

Description

Normalizes a numeric matrix of log-expression values by subtracting the median of each column.

Usage

median_centering(df, LOG2.names)

Arguments

Value

A numeric matrix of the same dimensions, now median-centered.

Examples

# Create a small, synthetic log-expression matrix
log_df <- data.frame(
    Sample1 = c(10.0, 10.2, 9.8),
    Sample2 = c(12.0, 12.1, 11.9)
)
row.names(log_df) <- c("ProtA", "ProtB", "ProtC")

LOG2.names <- c("Sample1", "Sample2")

normalized_df <- median_centering(log_df, LOG2.names)

print(normalized_df)

Description

Creates a heatmap using the base R heatmap function.

Usage

my_heatmap(data, cond.names, title)

Arguments

Value

Returns (invisibly) a list with components from the heatmap function, such as rowInd and colInd.

Examples

# Create synthetic data
heat_data <- data.frame(
    Protein = paste0("Prot", 1:5),
    Sample1 = rnorm(5, 10),
    Sample2 = rnorm(5, 12)
)
sample_names <- c("Sample1", "Sample2")

# Generate the plot
my_heatmap(heat_data, sample_names, "My Heatmap")

Description

Creates a ggplot2 heatmap of differentially expressed proteins.

Usage

my_heatmap_differential(limma, data, cond.names, title)

Arguments

Value

A ggplot object.

Examples

# Create synthetic data
limma_results <- data.frame(
    Protein = c("Prot1", "Prot2", "Prot3"),
    expression = c("Up-regulated", "Not Significant", "Down-regulated"),
    logFC = c(2.5, 0.5, -2.1)
)
data_expr <- data.frame(
    Protein = c("Prot1", "Prot2", "Prot3"),
    Sample1 = c(12, 10, 8),
    Sample2 = c(13, 10, 7)
)
sample_names <- c("Sample1", "Sample2")

# Generate the plot
my_heatmap_differential(limma_results, data_expr, sample_names, "DE Heatmap")

Description

A wrapper function to apply various normalization methods from the wrMisc package to the specified LOG2 intensity columns.

Usage

normalization_func(df, LOG2.names, method)

Arguments

Value

The original data frame (df) with the specified LOG2.names columns replaced by their normalized versions.

Examples

# Create synthetic LOG2 data
log_df <- data.frame(
    Protein = c("ProtA", "ProtB"),
    LOG2.Sample1 = c(10, 12),
    LOG2.Sample2 = c(11, 13)
)

# Specify which columns to normalize
log_cols <- c("LOG2.Sample1", "LOG2.Sample2")

# Run median normalization
normalized_df <- normalization_func(log_df, log_cols, "median")

print("Original Data:")
print(log_df)
print("Normalized Data:")
print(normalized_df)

Description

A helper function to find all column names containing "LOG2".

Usage

obtain_LOG.names(df)

Arguments

Value

A character vector of column names.

Examples

# Create a small example data frame
my_df <- data.frame(
    Protein = c("P1", "P2"),
    Intensity.A = c(100, 200),
    LOG2.A = c(6.6, 7.6),
    Intensity.B = c(110, 210),
    LOG2.B = c(6.7, 7.7)
)

# This example is runnable!
log_names <- obtain_LOG.names(my_df)
print(log_names)

Description

Identifies proteins that are uniquely present in one of two conditions, based on the presence of finite (e.g., LOG2 intensity) vs. non-finite (e.g., Inf, NA) values.

Usage

obtain_unique_proteins(df, metadata, selected_conditions)

Arguments

Details

This function is designed for label-free data where a protein's absence is marked by a non-finite value (like Inf from a log-transformation of 0). It identifies proteins present in at least half of the replicates of one condition while being completely absent (non-finite) in the other.

Value

A list containing two data frames: 1. cond1_unicas: A data frame of proteins unique to the first group. 2. cond2_unicas: A data frame of proteins unique to the second group.

Examples

# Create synthetic data with infinite values (representing missing)
df_prot <- data.frame(
    Protein = paste0("Prot", 1:4),
    LOG2.C1 = c(10, 12, Inf, 9),
    LOG2.C2 = c(11, 13, Inf, 10),
    LOG2.T1 = c(Inf, Inf, 8, 11),
    LOG2.T2 = c(Inf, Inf, 7, 12)
)
# Prot1/2 are present in Control, absent in Treatment
# Prot3 is present in Treatment, absent in Control
# Prot4 is present in both

metadata <- data.frame(
    log2_col = c("LOG2.C1", "LOG2.C2", "LOG2.T1", "LOG2.T2"),
    group = c("Control", "Control", "Treatment", "Treatment")
)
# Note: T is [1] (second_group), C is [2] (first_group)
selected_cond <- c("Treatment", "Control")

unique_list <- obtain_unique_proteins(df_prot, metadata, selected_cond)

# cond1_unicas (Control) should have Prot1, Prot2
print("Unique to Control:")
print(unique_list[[1]]$Protein)

# cond2_unicas (Treatment) should have Prot3
print("Unique to Treatment:")
print(unique_list[[2]]$Protein)

Description

Performs PCA and generates a 2D scatter plot of the first two principal components using base R plotting.

Usage

pca(x, metadata, selected_conditions, pc_x = 1, pc_y = 2)

Arguments

Value

This function is called for its side effect of generating a base R plot. It does not return an object.

Examples

# Create synthetic data
x_data <- data.frame(
    LOG2.C1 = rnorm(10, 8), LOG2.C2 = rnorm(10, 8),
    LOG2.T1 = rnorm(10, 9), LOG2.T2 = rnorm(10, 9),
    row.names = paste0("Prot", 1:10)
)
metadata <- data.frame(
    log2_col = c("LOG2.C1", "LOG2.C2", "LOG2.T1", "LOG2.T2"),
    group = c("Control", "Control", "Treatment", "Treatment")
)
selected_cond <- c("Treatment", "Control") # Matches user's logic [2] then [1]

# Generate the plot
pca(x_data, metadata, selected_cond)

Description

Generates a plot of the squared Coefficient of Variation (CV) versus the mean abundance (A). This is often used to check imputation quality.

Usage

plotCV2(y, trend = TRUE, main = "Imputation check", ...)

Arguments

Value

A data frame containing the mean (A) and CV (CV^2) values that were plotted.

Examples

# Create synthetic LOG2 intensity data
log_data <- matrix(rnorm(100, mean = 10, sd = 1), nrow = 20, ncol = 5)
log_data[sample(1:100, 10)] <- NA # Add some NAs

# Create the CV plot
cv_data <- plotCV2(log_data, main = "CV vs. Mean Plot")
print(head(cv_data))

Description

Plots the density of intensity values before and after imputation to visualize the effect of the imputation.

Usage

postimputation_state(data_filtered, imputation, LOG2.names)

Arguments

Value

A ggplot object showing the two density plots.

Examples

# This function depends on `impute_KNN_data` and `impute_data`
# We must create those functions and the data they need.

# 1. Create synthetic data
data_filt <- data.frame(
    Protein = c("A", "B", "C", "D"),
    LOG2.S1 = c(10, 12, NA, 8),
    LOG2.S2 = c(11, 13, 9, NA),
    KEEP = c(TRUE, TRUE, TRUE, TRUE) # For impute_data
)

# 2. Create minimal versions of the imputation functions for the example
impute_KNN_data <- function(df, lnames, k) {
    df[lnames][is.na(df[lnames])] <- mean(unlist(df[lnames]), na.rm = TRUE)
    return(df)
}
impute_data <- function(df, lnames) {
    df[lnames][is.na(df[lnames])] <- rnorm(sum(is.na(df[lnames])), mean = 7, sd = 0.3)
    return(df)
}

# 3. Run the plot function
postimputation_state(data_filt, imputation = 2, LOG2.names = c("LOG2.S1", "LOG2.S2"))

Description

Generates a missing value aggregation plot from the VIM package to visualize patterns of missing data.

Usage

preimputation_state(df, cond.names)

Arguments

Value

A plot showing the aggregation of missing values.

Examples

# Create synthetic data with NAs
log_df <- data.frame(
    LOG2.C1 = c(10, 10.1, NA, 9.9),
    LOG2.C2 = c(11, NA, NA, 11.2),
    LOG2.T1 = c(12, 12.1, 12.2, 12.3)
)

# Plot the missingness pattern
preimputation_state(log_df, c("LOG2.C1", "LOG2.C2", "LOG2.T1"))

Description

Generates a Q-Q plot to compare the distributions of two columns in a data frame.

Usage

qqplot_function(df, colname1, colname2, color)

Arguments

Value

A ggplot object.

Examples

my_data <- data.frame(
    Theoretical = rnorm(100),
    Sample = rnorm(100, mean = 0.5)
)
qqplot_function(my_data, "Theoretical", "Sample", "blue")

Description

Filters raw proteomics data (e.g., from MaxQuant) to remove contaminants, reverse hits, and site-only identifications. It also parses protein identifiers from Fasta headers.

Usage

quick_filtering(
  raw,
  platform,
  organism,
  metadata,
  selected_conditions,
  directory_path = NULL
)

Arguments

Value

A data frame filtered to remove contaminants/reverse hits, with new "Protein" and "Protein_description" columns.

Examples

raw_data <- data.frame(
    Potential.contaminant = c("", "+", "", ""),
    Reverse = c("", "", "+", ""),
    Only.identified.by.site = c("", "", "", "+"),
    Fasta.headers = c(
        ">sp|P04637|P53_HUMAN Cellular tumor antigen p53 OS=Homo sapiens",
        ">sp|P00000|CONTAM_Protein Common contaminant",
        ">sp|P12345|REV_Protein Reverse Sequence",
        ">sp|P62258|1433E_HUMAN 14-3-3 protein epsilon"
    ),
    # Intensity columns
    Intensity.S1 = c(10000, 500, 200, 100),
    Intensity.S2 = c(12000, 600, 150, 90),
    stringsAsFactors = FALSE
)

metadata <- data.frame(
    intensity_sample_name = c("Intensity.S1", "Intensity.S2"),
    sample_name = c("Sample_A", "Sample_B")
)

filtered_df <- quick_filtering(
    raw = raw_data,
    platform = 1,
    organism = 2,
    metadata = metadata,
    selected_conditions = NULL,
    directory_path = NULL
)

Description

This function builds and returns the Shiny app object for TraianProt.

Usage

runTraianProt()

Value

A shiny.appobj object representing the TraianProt app.

Examples

app <- runTraianProt()

if (interactive()) {
    shiny::runApp(app)
}

Description

Uses ggplot2 and ggExtra to create a scatterplot comparing two columns, with marginal density plots.

Usage

scatterplot_function(df, colname1, colname2)

Arguments

Value

A ggExtra plot object.

Examples

my_data <- data.frame(
    Sample1 = rnorm(100),
    Sample2 = rnorm(100)
)
scatterplot_function(my_data, "Sample1", "Sample2")

Description

A comprehensive function to perform differential expression (DE) analysis between two conditions. It supports t-tests, Wilcoxon tests, and limma, with options for paired data and DEqMS spectral count weighting.

Usage

statistical_analysis(
  df,
  test,
  paired = FALSE,
  metadata,
  logfc,
  sig,
  adjval,
  statval,
  unique_proteins,
  way,
  psms,
  platform,
  selected_conditions,
  diann_dir = NULL
)

Arguments

Details

Perform Differential Expression Statistical Analysis

This function acts as a wrapper for various statistical tests. The logFC is calculated as mean(treatment) - mean(control).

When way = 2, unique proteins are added to the results with imputed logFC and p-values (e.g., p-value = min(p.value)/100) to ensure they appear in visualizations.

The function uses several non-standard variable assignment methods (e.g., assign, ls, get) to dynamically create sample groups.

Value

A data frame (results.eb) with DE results, including logFC, p.value, adj.P.Val, expression, Protein, and Protein_description. Additional columns may be present for limma/DEqMS (e.g., sca.P.Value).

Examples

# 1. Create a mock data frame (df)
set.seed(123)
df_data <- data.frame(
    Protein = paste0("Prot", 1:5),
    Protein_description = paste("Description for", paste0("Prot", 1:5)),
    Cond1_R1 = rnorm(5, 10, 1),
    Cond1_R2 = rnorm(5, 10, 1),
    Cond2_R1 = rnorm(5, 12, 1), # Cond2 is higher
    Cond2_R2 = rnorm(5, 12, 1)
)
rownames(df_data) <- df_data$Protein

# 2. Create mock metadata
meta <- data.frame(
    log2_col = c("Cond1_R1", "Cond1_R2", "Cond2_R1", "Cond2_R2"),
    group = c("Condition1", "Condition1", "Condition2", "Condition2")
)

# 3. Define unique proteins (for way = 2)
uniques_list <- list(
    data.frame(Protein = "Prot_Control", Protein_description = "Only in Control"),
    data.frame(Protein = "Prot_Treat", Protein_description = "Only in Treatment")
)

# Run analysis (t-test, unpaired, no PSMs, include uniques)
results <- statistical_analysis(
    df = df_data,
    test = 1, # t-test
    paired = FALSE,
    metadata = meta,
    logfc = 1,
    sig = 0.05,
    adjval = "BH",
    statval = 2, # Use adjusted p-value
    unique_proteins = uniques_list,
    way = 2, # Include uniques
    psms = FALSE,
    platform = 1,
    selected_conditions = c("Condition2", "Condition1") # Treat 2 vs 1
)

print(results)

Description

Compiles the raw, normalized, and statistical results from the TraianProt pipeline into a Bioconductor-standard SummarizedExperiment object.

Usage

traian_to_SE(df.F, total_dataset, metadata, df_limma)

Arguments

Value

a Bioconductor-standard SummarizedExperiment object, containing the raw, normalized and statistical resutls.

Examples

df_raw <- data.frame(
    Protein = paste0("Protein_", 1:5),
    WT_1 = runif(5, 100, 1000), WT_2 = runif(5, 100, 1000),
    Treat_1 = runif(5, 100, 1000), Treat_2 = runif(5, 100, 1000)
)

df_norm <- df_raw
df_norm[, 2:5] <- log2(df_norm[, 2:5]) # Simulamos la normalización

meta <- data.frame(
    Condition = c("WT", "WT", "Treatment", "Treatment"),
    log2_col = c("WT_1", "WT_2", "Treat_1", "Treat_2")
)

stats <- data.frame(
    Protein = paste0("Protein_", 1:5),
    logFC = c(1.2, -0.5, 2.1, 0.1, -1.8),
    p_value = c(0.01, 0.4, 0.005, 0.8, 0.02)
)

se_object <- traian_to_SE(df_raw, df_norm, meta, stats)
se_object

Description

Calculates the theoretical power curve based on the observed variability in the experimental data and a desired Fold Change.

Usage

traianprot_power_curve(
  df,
  log2_cols,
  foldchange,
  replicatespower,
  alpha_level_choice,
  alpha_level
)

Arguments

Value

A ggplot object containing the power curve.

Examples

# Simulated data
data <- data.frame(
    ID = letters[1:10],
    R1 = rnorm(10), R2 = rnorm(10), R3 = rnorm(10)
)
plot <- traianprot_power_curve(
    data, c("R1", "R2", "R3"),
    foldchange = 2,
    replicatespower = 5,
    alpha_level_choice = 1,
    alpha_level = 0.05
)
print(plot)

Description

Generates a t-SNE plot for sample clustering using ggplot2.

Usage

tsne(x, metadata, perplexity_num, selected_conditions)

Arguments

Value

A ggplot object.

Examples

df <- data.frame(
    # Group A (Low values)
    Sample1 = c(10.1, 10.5, 12.0, 10.0),
    Sample2 = c(10.2, 10.6, 12.1, 10.1),
    Sample3 = c(10.3, 10.7, 12.2, 10.2),
    Sample4 = c(10.4, 10.8, 12.3, 10.3),
    Sample5 = c(10.5, 10.9, 12.4, 10.4),
    # Group B (High values)
    Sample6 = c(20.1, 21.5, 22.0, 20.0),
    Sample7 = c(20.2, 21.6, 22.1, 20.1),
    Sample8 = c(20.3, 21.7, 22.2, 20.2),
    Sample9 = c(20.4, 21.8, 22.3, 20.3),
    Sample10 = c(20.5, 21.9, 22.4, 20.4)
)
row.names(df) <- c("P1", "P2", "P3", "P4")

metadata <- data.frame(
    log2_col = paste0("Sample", 1:10),
    group = c(rep("A", 5), rep("B", 5))
)
selected_conditions <- c("A", "B")

if (requireNamespace("Rtsne") && requireNamespace("ggplot2")) {
    tsne_plot <- tsne(
        x = df,
        metadata = metadata,
        perplexity_num = 2,
        selected_conditions = selected_conditions
    )

    print(tsne_plot)
}

Description

Filters a protein data frame based on a minimum number of unique peptides found in a minimum fraction of replicates per condition.

Usage

unique_peptides_filter(df, metadata, number, min_fraction)

Arguments

Details

This function is useful for filtering data (e.g., from MaxQuant) to remove proteins with low confidence. It keeps a protein if, for at least one experimental group (condition), the unique peptide count is ⁠>= number⁠ in at least min_fraction of the samples in that group.

Value

A data frame (df_filtered) containing only the rows (proteins) that pass the filter.

Examples

# Create synthetic data
peptide_df <- data.frame(
    Pep.C1 = c(2, 0, 1, 5),
    Pep.C2 = c(3, 1, 0, 6),
    Pep.T1 = c(0, 2, 1, 0),
    Pep.T2 = c(0, 3, 0, 0)
)
rownames(peptide_df) <- c("ProtA", "ProtB", "ProtC", "ProtD")

meta <- data.frame(
    unique_peptides_col = c("Pep.C1", "Pep.C2", "Pep.T1", "Pep.T2"),
    group = c("Control", "Control", "Treatment", "Treatment")
)

# Filter: require min 1 peptide in at least 75% of reps (i.e., 2/2)
# ProtA: Passes in Control (2/2 have >= 1). Kept.
# ProtB: Passes in Treatment (2/2 have >= 1). Kept.
# ProtC: Fails in Control (1/2), Fails in Treatment (1/2). Dropped.
# ProtD: Passes in Control (2/2 have >= 1). Kept.

filtered_df <- unique_peptides_filter(
    df = peptide_df,
    metadata = meta,
    number = 1,
    min_fraction = 0.75
)

print("Filtered proteins (should be A, B, D):")
print(rownames(filtered_df))

Description

Generates a Venn diagram plot object from a list of unique proteins and a data frame of common proteins.

Usage

venn_diagram(df, unique_proteins, color1, color2)

Arguments

Value

A gList object (a plot) representing the Venn diagram.

Examples

common_proteins <- data.frame(
    Protein = c("ProteinA", "ProteinB", "ProteinC")
)

unique_protein_list <- list(
    Control = data.frame(Protein = c("ProteinD", "ProteinE")),
    Treatment = data.frame(Protein = c("ProteinF", "ProteinG", "ProteinH", "ProteinI"))
)

my_venn_plot <- venn_diagram(
    df = common_proteins,
    unique_proteins = unique_protein_list,
    color1 = "blue",
    color2 = "red"
)

grid::grid.newpage()
grid::grid.draw(my_venn_plot)

Description

This function generates an interactive volcano plot using the 'plotly' package based on differential expression results.

Usage

volcano_plot(limma, title, label, statval, psms)

Arguments

Details

The function selects the appropriate p-value column based on the statval and psms arguments. The y-axis label is automatically set to "-log10 p-value" or "-log10 q-value" (for adjusted p-values). Points are colored based on the expression column.

Value

A plotly object, which renders as an interactive volcano plot.

Examples

# Create a mock data frame for demonstration
limma_results <- data.frame(
    logFC = rnorm(100, 0, 2),
    Protein = paste0("PROT", 1:100),
    p.value = runif(100, 0, 1),
    adj.P.Val = runif(100, 0, 1),
    sca.P.Value = runif(100, 0, 1),
    sca.adj.pval = runif(100, 0, 1),
    expression = sample(c("Up-regulated", "Down-regulated", "Unchanged"),
        100,
        replace = TRUE
    )
)

# Generate an interactive plot using adjusted p-values without PSM weighting
volcano_plot(
    limma = limma_results,
    title = "Interactive Volcano Plot",
    label = 5,
    statval = 2,
    psms = FALSE
)

Description

This function generates a static volcano plot using the 'ggplot2' package, suitable for saving as a high-resolution image (e.g., TIFF) for publication.

Usage

volcano_plot_tiff(limma, title, label, statval, psms)

Arguments

Details

The function uses ggplot2::theme_light() and customizes axis text and line sizes. The y-axis label is automatically set to "-log10 p-value" or "-log10 q-value" (for adjusted p-values).

Value

A ggplot object, which renders as a static volcano plot.

Examples

mock_fc <- rnorm(100, 0, 2)

limma_results <- data.frame(
    logFC = mock_fc,
    log2FC = mock_fc, # Use the vector we just created
    Protein = paste0("PROT", 1:100),
    p.value = runif(100, 0, 1),
    adj.P.Val = runif(100, 0, 1),
    sca.P.Value = runif(100, 0, 1),
    sca.adj.pval = runif(100, 0, 1),
    expression = sample(c("Up-regulated", "Down-regulated", "Unchanged"),
        100,
        replace = TRUE
    )
)

# Generate a static plot
static_plot <- volcano_plot_tiff(
    limma = limma_results,
    title = "Static Volcano Plot",
    label = 3,
    statval = 2,
    psms = TRUE
)

print(static_plot)