Package 'TiDEomics'

Description

Calculate per-feature temporal properties for each group from the mean of replicates at each time point.

Usage

calc_feature_property(se_obj_merged_list, threshold = NULL)

Arguments

Value

A list of SummarizedExperiment objects, each corresponding to one group, with updated rowData containing the following columns: Feature, Group, Exp_threshold, T_exp (number of time points with values > Exp_threshold), T_total (total number of time points), P_trend (p-value from Bartels' test for randomness against a trend), Max_FC (maximum fold change across time points), Max_FC_time (difference between the time points at which maximum and minimum expression occur; positive if max occurs after min, negative otherwise), Exp_ratio (T_exp / T_total), Rho_time (Spearman correlation with time; positive = up trend), Peak_ratio (number of local maxima / T_total; 0 = no peaks detected), Log2_CV (log2(1 + CV) of expression across time points), AUC (area under the time-0-normalised curve from assay 2; NA if assay 2 is not available).

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)
example_obj_list <- split_groups(example_obj)
example_obj_merged_list <- merge_replicates(example_obj_list)

example_obj_merged_list <-
    calc_feature_property(example_obj_merged_list, threshold = 0)

Description

Calculate mean and standard deviation for each group and time point

Usage

calc_mean_sd(se_obj)

Arguments

Value

A list containing two data frames: one for original values and one for normalized values (if available). Each data frame includes columns for Mean, SD, Group, Time, and Feature.

Examples

data(example_obj)
table_mean_sd_list <- calc_mean_sd(example_obj)
table_mean_sd <- table_mean_sd_list$norm

Description

Check format of input data and sample annotation, create a SummarizedExperiment object

Usage

create_input(
  data,
  sample_ann,
  subject_col = NULL,
  replicate_col = NULL,
  batch_col = NULL
)

Arguments

Value

A SummarizedExperiment object containing the input data and sample annotations.

Examples

data <- data.frame(Feature = c("Gene1", "Gene2"),
                   Sample1 = c(1, 2), Sample2 = c(3, 4))
sample_ann <- data.frame(
    Sample = c("Sample1", "Sample2"),
    Group = c("A", "A"),
    Time = c(0, 1),
    Rep = c(1, 1),
    BatchID = c(1, 1)
)
se_obj <- create_input(data, sample_ann,
    replicate_col = "Rep", batch_col = "BatchID")

Description

Differential expression analysis between groups at each time point by limma. The function compares each pair of groups at each time point, and returns a nested list of DE analysis results for each pair of groups and each time point including all features, as well as a list of filtered DE results for each pair of groups based on the specified thresholds including only significant features. Use plot_DE_between_group() to visualise the number of DE features between groups over time.

Usage

DE_between_group(
  se_obj,
  group = NULL,
  filter = NULL,
  assay,
  adjP_thres = 0.05,
  logFC_thres = 1,
  trend = FALSE
)

Arguments

Details

Only time points present in both groups are compared. No multiple-testing correction is applied across the pairwise group-by-time comparisons; each comparison is independent. Apply your own correction (e.g., stats::p.adjust()) across combined results if needed.

Value

A list with: all_list (nested list of DE results per group pair and time point, all features); de_list (significant features only); fit_list (nested list of limma MArrayLM fit objects, for use with limma::plotSA()); ref_groups (group to be compared to); all_groups (all groups). The output can be passed to plot_DE_between_group() for visualisation.

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)

DE_between_group_out <- DE_between_group(example_obj, assay = 2)

Description

Differential expression analysis between time points within each group by limma

Usage

DE_between_time(
  se_obj,
  group = NULL,
  filter = NULL,
  assay,
  adjP_thres = 0.05,
  logFC_thres = 1,
  trend = FALSE
)

Arguments

Value

A list with: all_list (nested list of DE results per group and time comparison, all features); de_list (nested list of significant features only); fit_list (nested list of limma MArrayLM fit objects, for use with limma::plotSA()); time_series (vector of all available time points). The output can be passed to plot_DE_between_time() for visualisation.

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)

DE_between_time_out <- DE_between_time(example_obj, assay = 1)

Description

Variance decomposition by linear mixed models (LMM), to estimate the contribution of Group and Time to the variance of each feature. Modified from PALMO::lmeVariance() function.

Group and Time are always included as random effects. The formula is extended automatically based on colData:

• Subject column present: adds (1|Subject) to model between-subject differences

• interaction = TRUE: adds (1|Group:Time) (or (1|Subject:Time) when Subject present) to capture interaction variance. Requires >= 2 replicates per combination. Default: FALSE.

Output always includes: Group, Time, Residual. Subject is included when present. All values are percentages of total variance.

Allow using original data or data normalised to time point 0.

Usage

decomp_variance(
  se_obj,
  features = NULL,
  fixed_effect_var = NULL,
  interaction = FALSE,
  assay,
  core = 1
)

Arguments

Value

A data frame with variance decomposition results (percentages).

References

https://github.com/aifimmunology/PALMO/blob/main/R/lmeVariance.R

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)

var_decomp <- decomp_variance(example_obj, assay = 1)

Description

Enrichment analysis against MSigDB gene sets via hypergeometric test. Supports MSigDB categories, subcategories, and specified gene sets. Requires the msigdbr package.

Use cases:

• Drug perturbations: ⁠category = "C2", subcategory = "CGP"⁠ (chemical and genetic perturbations)

• TF targets: ⁠category = "C3", subcategory = "TFT:GTRD"⁠

• Hallmarks: ⁠category = "H", subcategory = NULL⁠ (no subcategory)

• GO: ⁠category = "C5", subcategory = "BP"⁠ (GO biological process)

Multiple testing: P-values are adjusted (Benjamini-Hochberg) per module across the gene-set tests within that module.

Usage

enrich_msigdb(
  gene_list,
  universe,
  category = NULL,
  subcategory = NULL,
  gene_sets = NULL,
  species = "Homo sapiens",
  db_species = "HS",
  name = NULL,
  pvalueCutoff = 0.05,
  pAdjustMethod = "BH",
  minGSSize = 10,
  maxGSSize = 500
)

Arguments

Value

A named list with one element: a data.frame with columns Cluster, ID, Description, GeneRatio, BgRatio, pvalue, p.adjust, Count, geneID.

Examples

if (requireNamespace("msigdbr", quietly = TRUE)) {
    data(example_net)
    example_module <- WGCNA_module(example_net) |>
        dplyr::filter(Module %in% c("1", "2"))
    hallmark_msigdb <- enrich_msigdb(example_module, category = "MH",
        species = "Mus musculus", db_species = "MM",
        minGSSize = 1,
        pvalueCutoff = 0.9,
        universe = WGCNA_module(example_net, exclude_grey = FALSE)$Feature)
}

Description

Gene ontology enrichment analysis of multiple gene sets with clusterProfiler, allowing for using different or same background genes for each gene set

Usage

enrichGO_list(
  gene_list,
  keyType = "SYMBOL",
  OrgDb,
  universe = NULL,
  universe_list = NULL,
  pAdjustMethod = "BH",
  pvalueCutoff = 0.05,
  qvalueCutoff = 0.05,
  category = NULL,
  simplify = FALSE,
  simplify_cutoff = 0.7,
  simplify_by = "p.adjust",
  simplify_select_fun = min,
  simplify_measure = "Wang",
  ...
)

Arguments

Value

A nested list of GO enrichment results with sublists: 'all' and 'simplified' (if simplify = TRUE) containing merged results of all or simplified terms across gene sets for each GO category; 'unmerged_all' and 'unmerged_simplified' (if simplify = TRUE) including all or simplified terms for each gene set and GO category.

Examples

if (requireNamespace("org.Mm.eg.db", quietly = TRUE)) {
    library(org.Mm.eg.db)
    library(clusterProfiler)
    data(example_net)
    # select two modules for demonstration
    example_module <- WGCNA_module(example_net) |>
        dplyr::filter(Module %in% c("1", "2"))
    # set cutoff to 1 to show all results for demonstration
    example_go_list = enrichGO_list(example_module, OrgDb = org.Mm.eg.db,
        universe = WGCNA_module(example_net, exclude_grey = FALSE)$Feature,
        pvalueCutoff = 1, qvalueCutoff = 1,
        category = "BP", simplify = FALSE)
}

Description

Gene ontology enrichment analysis of a ranked gene list with clusterProfiler. Genes can be ranked by variance decomposition results.

Usage

enrichGO_rank(
  rank_table,
  gene_rank_by,
  OrgDb,
  keyType = "SYMBOL",
  go_rank_by = "p.adjust",
  category = NULL,
  pvalueCutoff = 0.05,
  pAdjustMethod = "BH",
  ...
)

Arguments

Value

A list of gseaResult objects containing the GSEA results

Examples

if (requireNamespace("org.Mm.eg.db", quietly = TRUE)) {
    library(org.Mm.eg.db)
    data(example_obj)
    example_obj <- normalise_to_start(example_obj)

    var_decomp <- decomp_variance(example_obj,
        features = rownames(example_obj)[1:100], assay = 1)
    example_go_rank <- enrichGO_rank(var_decomp, gene_rank_by = "Time",
        OrgDb = org.Mm.eg.db, keyType = "SYMBOL",
        category = "BP")
}

Description

Enrichment analysis of multiple gene sets against databases from enrichR (e.g. DSigDB for drug signatures, ChEA for TF targets, KEGG for pathways, DrugBank for drug targets).

Usage

enrichR_list(
  gene_list,
  databases = c("DSigDB", "DrugMatrix"),
  site = "Enrichr",
  universe = NULL,
  universe_list = NULL,
  pvalueCutoff = 0.05,
  include_overlap = FALSE
)

Arguments

Details

Multiple testing: P-values are adjusted (Benjamini–Hochberg) independently for each module within each database. No correction is applied across databases or across modules.

Requires an internet connection.

Value

A named list of data.frames, one per database. Each data.frame has columns Cluster, Description, p.adjust (Adjusted.P.value from enrichR output), Combined.Score, Genes, and any additional columns returned by the enrichR API. Compatible with plot_modules_h(enrich_list = result, enrich_category = "DSigDB").

Examples

data(example_net)
library(dplyr)
example_module <- WGCNA_module(example_net) |>
    dplyr::filter(Module %in% c("1", "2"))
# Use high pvalueCutoff for demonstration
# enrichr_out <- enrichR_list(example_module,
#     databases = c("KEGG_2019_Mouse"),
#     universe = WGCNA_module(example_net, exclude_grey = FALSE)$Feature,
#     pvalueCutoff = 0.5)

Description

Code for producing the data is in inst/script/generate_example_go.R

Usage

data(example_go)

Format

A nested list with GO enrichment results for modules "1" and "2" from 'example_net', containing categories "BP" and "CC", with simplify = TRUE applied:

all

Merged results across gene sets for each GO category

simplified

Simplified merged results

unmerged_all

Per-gene-set results for each category

unmerged_simplified

Simplified per-gene-set results

Source

GSE263759

Description

Code for producing the data is in inst/script/generate_example_net.R

Usage

data(example_net)

Format

A list including WGCNA module assignments, module eigengenes, dendrogram, input data, sample information, and parameters used

Source

GSE263759

Description

A subset of data_obj <- create_input(data = tutorial_data, sample_ann = tutorial_sample_info) for use in runnable examples in function documentation.

Usage

data(example_obj)

Format

A SummarizedExperiment object with assays of 100 rows and 40 columns, colData of 40 rows and 5 columns:

colData

tutorial_sample_info

assays

tutorial_data first 100 rows, "Feature" column as rownames

Details

Code for preparing the data is available in inst/script/tutorial_input.R

Source

GSE263759

Description

Code for preparing the data is in inst/script/generate_example_res_list.R

Usage

data(example_res_list)

Format

A nested list, each element is a list with Trendy results for one group

Source

GSE263759

Description

Identifies hub features within each module using module membership (kME). For unsigned networks, absolute kME values are used to reproduce the behaviour of WGCNA::blockwiseModules(). For signed and signed hybrid networks, signed kME values are retained.

Usage

extract_hubs(net, top_n = 3, exclude_grey = TRUE)

Arguments

Value

A character vector containing the top top_n features from each module concatenated together.

Examples

data(example_net)
hubs <- extract_hubs(example_net, top_n = 3)

Description

Extract features with different segment trends from trendy segmented regression results summary

Usage

extract_segment_trends(trendy.summary)

Arguments

Value

A nested list of features grouped by their segment trend patterns within each group.

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)
example_obj_list <- split_groups(example_obj)
example_obj_merged_list <- merge_replicates(example_obj_list)
example_obj_merged_list <- calc_feature_property(example_obj_merged_list,
    threshold = 0)
# no missing value in the example dataset, so imputation is not necessary
example_obj_merged_imp_list <- impute_groups(example_obj_merged_list)

data("example_res_list")
trendy_summary <- summarise_Trendy(example_res_list)
trendy_list <- extract_segment_trends(trendy_summary)

Description

Flattens the nested DE result structure returned by DE_between_group() or DE_between_time() into a named list of data frames, one entry per contrast-time combination. This is useful for exporting results to other tools such as DeeDeeExperiment.

Usage

flatten_DE(de_list)

Arguments

Value

A named list of data frames with columns renamed for DeeDeeExperiment compatibility (log2FoldChange, pvalue, padj). Returns an empty list if de_list is empty.

Examples

data(tutorial_data)
data(tutorial_sample_info)
tide <- prepare_tide(tutorial_data, tutorial_sample_info,
    keep = "threshold", residual_threshold = 100)
tide$DE <- DE_between_group(tide$se, assay = "norm",
    filter = 1, trend = TRUE)
de_flat <- flatten_DE(tide$DE)
str(de_flat, max.level = 1)

Description

Flattens the nested enrichment result structure returned by enrichGO_list(), enrichR_list(), or enrich_msigdb() into a flat named list of data frames or enrichResult objects. This is useful for exporting results to other tools such as DeeDeeExperiment.

Usage

flatten_enrich(enrich_list)

Arguments

Value

A flat named list of data frames or enrichResult objects with DeeDeeExperiment-compatible names and columns (e.g. "clusterProfiler_BP_1", "enrichr_DSigDB"). Returns an empty list if enrich_list is empty.

Examples

data(example_go)
enrich_flat <- flatten_enrich(example_go)
str(enrich_flat, max.level = 1)

Description

Get the custom color palette set by set_custom_palette. If no custom palette has been set, the function will return a default color palette based on the number of groups specified in the groups argument.

Usage

get_custom_palette(groups)

Arguments

Value

A named vector of colors corresponding to the specified group names.

Examples

get_custom_palette(c("untreated", "IFNbeta"))

Description

Identify features that are unique to selected groups (present in those groups but not in any others) and annotate them with gene names using the clusterProfiler::bitr() function. The function also provides an option to perform Gene Ontology (GO) enrichment analysis on the identified unique features using the enrichGO_list() function and visualize the results with a dot plot.

Usage

group_specific_features(
  property_random_fc,
  groups = NULL,
  filter_ratio = 0.5,
  group_pct = 1,
  genename = TRUE,
  GO = TRUE,
  OrgDb = NULL,
  keytype = NULL,
  ...
)

Arguments

Value

A named list with element features (character vector of features identified as unique to the specified groups based on the filtering criteria). If genename is TRUE, element genename contains a data.frame of gene annotations. If GO is TRUE and enrichment succeeds, element GO contains a dot plot of GO enrichment results. Returns NULL if no unique features pass the filter.

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)
example_obj_list <- split_groups(example_obj)
example_obj_merged_list <- merge_replicates(example_obj_list)

example_obj_merged_list <-
    calc_feature_property(example_obj_merged_list, threshold = 0)
property_random_fc <- summarise_feature_property(example_obj_merged_list)

group_specific_features(property_random_fc, groups = c("untreated"),
    genename = FALSE, GO = FALSE)

Description

Impute missing values for each group of samples in a list of SummarizedExperiment objects. By default, replaces NA with the minimum non-NA value per group / subject. A custom function can be supplied for other strategies.

The input samples can contain replicates, or merged replicates (mean of replicates).

Usage

impute_groups(se_obj_list, fun = min, impute_by = c("group", "subject"))

Arguments

Value

A list of SummarizedExperiment objects with missing values imputed. Each object corresponds to one group of samples.

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)
example_obj_list <- split_groups(example_obj)
example_obj_merged_list <- merge_replicates(example_obj_list)
example_obj_merged_imp_list <- impute_groups(example_obj_merged_list)

Description

Merge SummarizedExperiment object of different groups into one

Usage

merge_groups(se_obj_list)

Arguments

Value

A SummarizedExperiment object containing all samples from the input list. The 'Group' column in the colData will indicate the group of each sample. New sample names will be prefixed with the group name.

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)
example_obj_list <- split_groups(example_obj)
example_obj_merged_list <- merge_replicates(example_obj_list)
example_obj_merged <- merge_groups(example_obj_merged_list)

Description

Calculate mean of replicates for each feature at each time point for each group.

When a Subject column is present, merging is done in two stages:

1. Average replicates within each Group-Subject-Time combination.

2. Average across subjects within each Group-Time combination. This gives equal weight to each subject regardless of replicate count.

Without a Subject column, all samples at the same Group-Time are averaged together in a single step.

Usage

merge_replicates(se_obj_list)

Arguments

Value

A list of SummarizedExperiment objects containing the mean of replicates for each feature at each time point for each group. Each object in the list corresponds to one group of samples.

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)
example_obj_list <- split_groups(example_obj)
example_obj_merged_list <- merge_replicates(example_obj_list)

Description

Normalise to starting time point, to make the mean of starting time point samples 0.

Two modes are available:

• by_subject = FALSE (default): computes the group-level mean at the first non-NA time point per feature and subtracts it. All samples within a group share the same baseline.

• by_subject = TRUE: computes each subject's value at the first non-NA time point per feature and subtracts that subject-specific baseline. Removes between-subject baseline differences. Use when each subject has their own initial condition. Requires a Subject column in colData.

Usage

normalise_to_start(se_obj, by_subject = FALSE)

Arguments

Value

A SummarizedExperiment object with normalised data in the second assay slot.

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)

Description

Plot breakpoint distribution among time points for each group

Usage

plot_breakpoints(res_list, group = NULL, fontsize = 8, ...)

Arguments

Value

A plot showing the distribution of breakpoints over time for each specified group

Examples

data("example_res_list")
plot_breakpoints(example_res_list)

Description

Plot correlation matrix between samples as a heatmap

Usage

plot_cor_matrix(
  se_obj,
  assay = 1,
  use = "pairwise.complete.obs",
  method = c("spearman", "pearson", "kendall"),
  label_group = TRUE,
  label_time = TRUE,
  label_rep = TRUE,
  label_batch = TRUE,
  show_rownames = FALSE,
  show_colnames = FALSE,
  fontsize = 8,
  cellwidth = 1,
  cellheight = 1,
  title = NULL,
  ...
)

Arguments

Value

A heatmap showing the correlation between samples.

Examples

data(example_obj)
plot_cor_matrix(example_obj)

Description

Plot the distribution of coefficient of variation (CV) for each feature with replicates. The CV is calculated as the standard deviation divided by the mean of the abundance values.

Note: CV is only meaningful for positive-valued data.

Usage

plot_cv(se_obj, fontsize = 8)

Arguments

Value

A plot showing the distribution of coefficient of variation (CV) for each feature, grouped by Time and Group. If there are no replicates, a message will be printed indicating that CV cannot be calculated.

Examples

data(example_obj)
plot_cv(example_obj)

Description

Plot the number of differentially expressed features between groups over time, using the output of DE_between_group().

DE features were pre-filtered with DE_between_group(), but by specifying adjP_thres and logFC_thres, users can re-filter the DE features for plotting.

Usage

plot_DE_between_group(
  DE_between_group_out,
  group = NULL,
  fontsize = 8,
  adjP_thres = NULL,
  logFC_thres = NULL
)

Arguments

Value

A series of line plots showing the number of DE features over time for each group comparison.

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)

DE_between_group_out <- DE_between_group(example_obj, assay = 2)
plot_DE_between_group(DE_between_group_out, fontsize = 8)

Description

Plot the number of differentially expressed features between time points within each group with heatmaps

DE features were pre-filtered with DE_between_time(), but by specifying adjP_thres and logFC_thres, users can re-filter the DE features for plotting.

Usage

plot_DE_between_time(
  DE_between_time_out,
  value = TRUE,
  fontsize = 8,
  nrow = 1,
  heatmap_width = 4,
  heatmap_unit = "cm",
  adjP_thres = NULL,
  logFC_thres = NULL
)

Arguments

Value

One heatmap per group showing the number of DE features between time points for each group, with the same scale across groups for easy comparison. The heatmap cells are optionally annotated with the number of DE features.

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)

DE_between_time_out <- DE_between_time(example_obj, assay = 1)
plot_DE_between_time(DE_between_time_out,
    fontsize = 8, value = TRUE, nrow = 1, heatmap_width = 3)

Description

This function generates density plots to visualise the distribution of abundance values across samples. The user can choose to facet the plot by "Group", "Time", or "Sample" to compare distributions across different conditions or samples. If no faceting variable is specified, a single density plot will be generated for all samples combined.

Usage

plot_distribution(se_obj, facet_by = NULL, fontsize = 8)

Arguments

Value

A plot showing the density distribution of abundance values, optionally faceted by the specified variable.

Examples

data(example_obj)
plot_distribution(example_obj)
plot_distribution(example_obj, facet_by = "Group")

Description

Visualization of GO enrichment analysis with dotplot, cnetplot, or emapplot in clusterProfiler

Usage

plot_GO(
  go_list,
  plot_dotplot = TRUE,
  plot_cnetplot = FALSE,
  plot_emapplot = FALSE,
  showCategory_dotplot = 5,
  showCategory_cnetplot = 5,
  showCategory_emapplot = 5,
  fontsize = 8,
  label = "genes",
  ...
)

Arguments

Value

A list of plots visualizing the GO enrichment results.

Examples

data(example_go)
plot_GO(example_go$all, plot_dotplot = TRUE,
    plot_emapplot = FALSE, plot_cnetplot = FALSE)

Description

Plot the number of identified features (i.e., features with non-missing values) for each sample

Usage

plot_ID(se_obj, fontsize = 8, signif = FALSE, ...)

Arguments

Value

A plot showing the ID number for each sample, with a dashed line indicating the average number across all samples. And a boxplot comparing the ID number between groups, with optional significance annotations.

Examples

# simulate data with random missing values
na_data <- matrix(rnorm(1500), nrow = 100, ncol = 150)
na_data[sample(length(na_data), size = 2000)] <- NA
na_data <- data.frame(Feature = paste0("Feature", 1:100), na_data)
colnames(na_data)[-1] <- paste0("Sample", 1:150)

na_obj <- create_input(na_data,
    data.frame(Sample = paste0("Sample", 1:150),
    Time = rep(rep(1:10, each = 5), 3),
    Group = rep(c("A", "B", "C"), each = 50),
    Replicate = rep(1:5, 30)))
plot_ID(na_obj, signif = TRUE)

Description

Plot the ratio of missing values for each sample

Usage

plot_missing(se_obj, fontsize = 8, signif = FALSE, ...)

Arguments

Value

A plot showing the missing value ratio for each sample, with a dashed line indicating the global missing value ratio across all samples. And a boxplot comparing the missing value ratio between groups, with optional significance annotations.

Examples

# simulate data with random missing values
na_data <- matrix(rnorm(1500), nrow = 100, ncol = 150)
na_data[sample(length(na_data), size = 2000)] <- NA
na_data <- data.frame(Feature = paste0("Feature", 1:100), na_data)
colnames(na_data)[-1] <- paste0("Sample", 1:150)

na_obj <- create_input(na_data,
    data.frame(Sample = paste0("Sample", 1:150),
    Time = rep(rep(1:10, each = 5), 3),
    Group = rep(c("A", "B", "C"), each = 50),
    Replicate = rep(1:5, 30)))
plot_missing(na_obj, signif = TRUE)

Description

Plot WGCNA modules' heatmaps, mean expression profiles, and enrichment terms, aligned horizontally.

Different from running WGCNA, the input data should have the replicates merged, instead of having multiple samples per group, time and feature (gene).

If certain time points are missing in some groups, NA values are added.

Usage

plot_modules_h(
  module,
  se_obj_merged,
  scale = TRUE,
  assay,
  ylabel = "Log2 abundance",
  profile_width = 3,
  profile_link_width = 1,
  enrich_list = NULL,
  enrich_category = "BP",
  enrich_rank_by = "p.adjust",
  enrich_top_n = 3,
  enrich_threshold = 0.05,
  fontsize = 8,
  heatmap_width = 8,
  heatmap_height = 8,
  mark_features = NULL,
  width = 16,
  height = 12,
  res = 300,
  suffix = "",
  device = "png",
  save = NULL
)

Arguments

Value

A combined plot of heatmap, mean expression profile, and enrichment terms for each WGCNA module

References

https://github.com/junjunlab/ClusterGVis

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)
example_obj_list <- split_groups(example_obj)
example_obj_merged_list <- merge_replicates(example_obj_list)
example_obj_merged <- merge_groups(example_obj_merged_list)

data(example_net)
# select two modules for demonstration
example_module <- WGCNA_module(example_net) |>
    dplyr::filter(Module %in% c("1", "2"))

data(example_go)

plot_modules_h(example_module |> dplyr::filter(Module != '0'),
    example_obj_merged, assay = 2, scale = TRUE,
    ylabel = "Z-score of log2 expression",
    enrich_list = example_go$all,
    enrich_category = "BP",
    heatmap_width = 6, heatmap_height = 4)

Description

Plot WGCNA modules' mean expression profiles and heatmaps, align vertically.

Different from running WGCNA, the input data should have the replicates merged, instead of having multiple samples per group, time and feature (gene).

If certain time points are missing in some groups, NA values are added.

Usage

plot_modules_v(
  module,
  se_obj_merged,
  scale = TRUE,
  assay,
  ylabel = "Log2 abundance",
  suffix = "",
  device = "png",
  save = NULL,
  width = 12,
  height = 8,
  height_ratio = 2,
  fontsize = 8,
  res = 300
)

Arguments

Value

Two plots aligned vertically by groups: the top one is a line plot of module feature mean expression profiles, the bottom one is a heatmap of feature expression across time points.

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)
example_obj_list <- split_groups(example_obj)
example_obj_merged_list <- merge_replicates(example_obj_list)
example_obj_merged <- merge_groups(example_obj_merged_list)

data(example_net)
example_module <- WGCNA_module(example_net)

plot_modules_v(example_module |> dplyr::filter(Module != '0'),
    example_obj_merged, assay = 2, scale = TRUE,
    ylabel = "Z-score of log2 (expression)",
    height_ratio = 2,
    fontsize = 6)

Description

Plot principal component analysis (PCA), using features without missing values

The samples are coloured by Group and sized by Time. Ellipses are drawn for each Group.

Usage

plot_pca(
  se_obj,
  plot_screeplot = TRUE,
  plot_loadings = TRUE,
  plot_morepc = TRUE,
  morepc = seq(1, 5),
  pc1 = 1,
  pc2 = 2,
  circle = FALSE,
  xlim_min = NULL,
  xlim_max = NULL,
  ylim_min = NULL,
  ylim_max = NULL,
  fontsize = 8,
  assay = 1
)

Arguments

Value

A PCA plot showing the distribution of samples, other plots provided by PCAtools package, and PCAtools output object for custom plotting.

Examples

data(example_obj)
PC = plot_pca(example_obj, morepc = seq(1, 3))

Description

Plot principal component analysis (PCA) results in 3D. This function takes the output PCA object of the plot_pca function and visualizes the samples in a 3D space defined by the specified principal components.

The samples are coloured by Group and sized by Time.

Usage

plot_pca_3D(pca, pcs = seq(1, 3))

Arguments

Value

An interactive 3D PCA plot showing the distribution of samples in the space defined by the specified principal components. The samples are coloured by Group and sized by Time.

Examples

data(example_obj)
PC = plot_pca(example_obj, morepc = seq(1, 3))
if (requireNamespace("plotly", quietly = TRUE)) {
    plot_pca_3D(PC$pca, pcs = seq(1, 3))
}

Description

Plot PCA with arrows indicating trajectory over time, for a single group.

Usage

plot_pca_arrows(
  se_obj,
  circle = TRUE,
  arrow = TRUE,
  pc1 = 1,
  pc2 = 2,
  fontsize = 8,
  assay = 1
)

Arguments

Value

A PCA plot coloured by Time, with optional ellipses and trajectory arrows.

Examples

data(example_obj)
plot_pca_arrows(example_obj[, example_obj$Group == "IFNbeta"])

Description

Plot PCA for each group separately, with optional circles around time points and arrows indicating trajectory over time. Accepts either a SummarizedExperiment object or a list of them (from split_groups()).

Usage

plot_pca_by_group(
  se_obj,
  circle = TRUE,
  arrow = TRUE,
  pc1 = 1,
  pc2 = 2,
  nrow = 1,
  fontsize = 8,
  assay = 1,
  legend_pos = "right"
)

Arguments

Value

A series of PCA plots showing the distribution of samples in each group, coloured by Time.

Examples

data(example_obj)
plot_pca_by_group(example_obj, circle = TRUE, arrow = TRUE)

# Also accepts a list from split_groups()
example_obj_list <- split_groups(example_obj)
plot_pca_by_group(example_obj_list)

Description

Plot segmented regression results for specified features in each group, using the Trendy::plotFeature() function. The input is a list of SummarizedExperiment objects with no missing values and a list of Trendy analysis results for each group.

Usage

plot_segments(
  se_obj_imp_list,
  res_list,
  feature,
  group = NULL,
  nrow = NULL,
  ...
)

Arguments

Value

Plots of the segmented regression fits for the specified features in each group.

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)
example_obj_list <- split_groups(example_obj)
example_obj_merged_list <- merge_replicates(example_obj_list)
example_obj_merged_list <-
    calc_feature_property(example_obj_merged_list, threshold = 0)
# no missing value in the example dataset, so imputation is not necessary
example_obj_merged_imp_list <- impute_groups(example_obj_merged_list)

data("example_res_list")
plot_segments(example_obj_merged_imp_list, example_res_list,
    feature = c("Kif3a"))

Description

Plot trend of feature abundances / expression over time by mean and standard deviation (SD) for each group. Accepts either a SummarizedExperiment object (computing mean/SD internally via calc_mean_sd()) or a pre-computed table from calc_mean_sd().

Usage

plot_trend(
  se_obj,
  assay = 1,
  groups = NULL,
  features,
  title = NULL,
  ylabel = "Log2 abundance",
  errorbar = TRUE,
  fontsize = 8
)

Arguments

Value

Plot of feature abundances over time by mean and SD

Examples

data(example_obj)
plot_trend(example_obj,
    features = sample(rownames(example_obj), 4))

Description

Plot UMAP of samples, using features without missing values

The UMAP plots are labelled by Group, Time, Replicate (if more than 1), Batch (if more than 1), and number of identified features (non-missing values).

Usage

plot_umap(
  se_obj,
  seed = 1234,
  plot_ID = FALSE,
  circle = FALSE,
  xlim_min = NULL,
  xlim_max = NULL,
  ylim_min = NULL,
  ylim_max = NULL,
  umap_neighbors = NULL,
  fontsize = 8,
  assay = 1
)

Arguments

Value

A list of UMAP plots showing the distribution of samples. And a data frame containing UMAP coordinates and sample annotations for custom plotting.

Examples

data(example_obj)
umap_layout <- plot_umap(example_obj)

Description

Plot UMAP for each group separately. Accepts either a SummarizedExperiment object or a named list of them (from split_groups()).

Usage

plot_umap_by_group(
  se_obj,
  seed = 1234,
  nrow = 1,
  umap_neighbors = NULL,
  fontsize = 8,
  assay = 1,
  legend_pos = "right"
)

Arguments

Value

A series of UMAP plots showing the distribution of samples in each group, coloured by Time.

Examples

data(example_obj)
plot_umap_by_group(example_obj)

# Also accepts a list from split_groups()
example_obj_list <- split_groups(example_obj)
plot_umap_by_group(example_obj_list)

Description

Based on PALMO::variancefeaturePlot() function

Usage

plot_variance(
  var_decomp,
  rank = "Group",
  features = NULL,
  top_n = 20,
  show_ylab = TRUE,
  fontsize = 8
)

Arguments

Value

A stacked bar plot showing the percentage of variance explained by each model component (Group, Time, Residual, plus Subject and interaction terms when present). Features are ranked by the specified rank variable (Group or Time) and only the top n features are plotted if features is not specified.

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)

var_decomp <- decomp_variance(example_obj, assay = 1)
plot_variance(var_decomp, rank = "Time", top_n = 20)

Description

This function creates a volcano plot to visualize the results of differential expression (DE) analysis performed by DE_between_group() or DE_between_time(). The plot displays log2 fold change on the x-axis and -log10 adjusted p-value on the y-axis, with DE features highlighted based on specified thresholds. If no thresholds are provided, the function will use the thresholds that were applied when running DE_between_group() or DE_between_time().

Usage

plot_volcano(
  DE_out,
  group1,
  group2,
  time,
  group,
  time1,
  time2,
  logFC_thres = NULL,
  adjP_thres = NULL,
  label = FALSE,
  fontsize = 8,
  ...
)

Arguments

Value

A volcano plot of DE results

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)

DE_between_group_out <- DE_between_group(example_obj, assay = 2)
plot_volcano(DE_between_group_out, group1 = "untreated",
    group2 = "IFNbeta", time = 24,
    logFC_thres = 0.5, adjP_thres = 0.05, label = TRUE)

Description

Plot WGCNA module eigengenes and module-group/time correlations based on output of run_WGCNA()

Usage

plot_WGCNA(net, fontsize = 8)

Arguments

Value

Plots of WGCNA module dendrogram, module eigengenes, pairwise scatterplots of eigengenes, clustering of module eigengenes, and module-trait correlation heatmap

References

https://github.com/edo98811/WGCNA_official_documentation/blob/main/FemaleLiver-03-relateModsToExt.R

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)

data("example_net")
plot_WGCNA(example_net, fontsize = 8)

Description

One-step data preparation wrapper for TiDEomics: creates the SummarizedExperiment object, normalises to starting time point, filters out group-specific features, runs variance decomposition, and filters out noisy features with high residual variance. Returns both unmerged and merged (mean of replicates) outputs, with all-feature and filtered versions of each, ready for downstream analysis.

Usage

prepare_tide(
  data,
  sample_ann,
  subject_col = NULL,
  replicate_col = NULL,
  batch_col = NULL,
  feature_threshold = 0,
  filter_ratio = 0.5,
  min_groups = 2,
  keep = c("below_quantile", "top_n", "threshold"),
  residual_threshold = NULL,
  n_keep = 5000,
  interaction = FALSE,
  n_cores = 1
)

Arguments

Details

The pipeline:

1. create_input(): validate and build SE

2. normalise_to_start(): add time-0-normalised assay

3. split_groups() -> merge_replicates() -> calc_feature_property() -> summarise_feature_property(): per-feature expression statistics

4. group_specific_features(): filter out group-specific features

5. decomp_variance() : LMM variance decomposition on both "orig" and "norm" assays

6. Residual filter: filter out features with high residual variance, calculated from decomp_variance() on each assay

7. Produce filtered-unmerged and filtered-merged SEs for each assay

Value

A named list with elements. The elements suffixed with "_filtered" have passed both the cross-group and residual filters, and are split into $orig and $norm sub-lists, with variance decomposition and residual filtering done per assay.

1. se (SE with replicates, all features, both assays);

2. se_filtered (a list of two SEs with replicates, filtered features by orig / norm, use for WGCNA);

3. merged_list (per-group merged SEs, all features, use for Trendy);

4. merged_list_filtered (a list of two per-group merged SEs, filtered features by orig / norm);

5. merged_se (single merged SE, all groups combined, all features, use for plot_modules_v/h);

6. merged_se_filtered (a list of two single merged SEs, filtered features by orig / norm, use for plot_modules_v/h);

7. variance (a list of two variance decomposition data.frames, computed from orig / norm assay);

8. feature_property (feature property summary, all features);

9. filter_summary (per-stage filtering statistics);

10. group_specific_filter (a vector of group-specific features, removed by the cross-group filter); 11-13. DE, enrichment, WGCNA (empty on initialization).

Examples

data(tutorial_data)
data(tutorial_sample_info)
tide <- prepare_tide(tutorial_data, tutorial_sample_info,
    keep = "threshold", residual_threshold = 100)
tide$filter_summary

Description

Prepare input data for WGCNA (format, QC), then choose the appropriate soft thresholding power for network construction, by analysing scale free topology with different soft thresholding powers.

Usage

prepare_WGCNA(
  se_obj,
  assay,
  networkType = "signed",
  RsquaredCut = 0.8,
  MeanConnectivity = 100,
  powers = NULL,
  fontsize = 8,
  ...
)

Arguments

Value

A list containing results of the scale-free topology fit indices with different powers, suggested power, network type and prepared input data used for reuse in run_WGCNA()

References

https://github.com/edo98811/WGCNA_official_documentation/

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)

wgcna_input <- prepare_WGCNA(example_obj, assay = 2, powers = seq(1, 30),
    networkType = "signed", RsquaredCut = 0.8)
wgcna_input$fitIndices
picked_power <- wgcna_input$powerEstimate

Description

Run segmented regression analysis with Trendy on imputed data for multiple groups of samples in a list of SummarizedExperiment objects. The results will be returned in a list format. Each element in the list corresponds to one group of samples and contains the Trendy analysis results for that group.

If the feature parameter is not specified, all features expressed in at least minExp time points in each group will be included in the analysis for the group. calc_feature_property() should be run before imputation to calculate the expression ratio for each feature in each group. Missing values can be imputed after calc_feature_property() using the impute_groups() function.

If a group has less than 4 time points are available, Trendy analysis cannot be performed and NULL will be returned for the group.

Usage

run_Trendy(
  se_obj_imp_list,
  group = NULL,
  feature = NULL,
  minExp = 0.5,
  maxK = 1,
  meanCut = 0,
  minNumInSeg = 3,
  NCores = 1,
  ...
)

Arguments

Value

A list of Trendy analysis results, including the fitted model parameters and statistics for each feature in each group.

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)
example_obj_list <- split_groups(example_obj)
example_obj_merged_list <- merge_replicates(example_obj_list)
example_obj_merged_list <- calc_feature_property(example_obj_merged_list,
    threshold = 0)
# no missing value in the example dataset, so imputation is not necessary
example_obj_merged_imp_list <- impute_groups(example_obj_merged_list)

# "untreated" group has only 3 time points, so Trendy analysis will not be
# performed for this group
example_res_list <- run_Trendy(example_obj_merged_imp_list, maxK = 1,
    minNumInSeg = 2, meanCut = 0,
    feature = rownames(example_obj_merged_imp_list[[1]])[1:10])

Description

Run WGCNA based on output of prepare_WGCNA(), to identify co-expression modules of features with WGCNA::blockwiseModules()

Usage

run_WGCNA(wgcna_input, power, numericLabels = TRUE, ...)

Arguments

Value

The built network and parameters of WGCNA::blockwiseModules(), and the input data and sample information for use in plot_WGCNA().

References

https://github.com/edo98811/WGCNA_official_documentation/blob/main/FemaleLiver-03-relateModsToExt.R

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)

wgcna_input <- prepare_WGCNA(example_obj, assay = 2, powers = seq(1, 30),
    networkType = "signed", RsquaredCut = 0.8)
wgcna_input$fitIndices
picked_power <- wgcna_input$powerEstimate
example_net <- run_WGCNA(wgcna_input,
    power = picked_power,
    minModuleSize = 10, # only 100 genes in the example data
    numericLabels = TRUE)

Description

Set a custom color palette for groups to use in all subsequent plotting functions that support custom palettes.

Usage

set_custom_palette(palette)

Arguments

Value

The function does not return a value but sets the custom palette as an option that can be accessed by other plotting functions. The palette will be stored in the R session options under the name "custom.palette".

Examples

custom_palette <- c("untreated" = "#1b9e77", "IFNbeta" = "#d95f02",
    "IFNgamma" = "#7570b3", "LPS" = "#e7298a")
set_custom_palette(custom_palette)

Description

Split SummarizedExperiment object by groups

Usage

split_groups(se_obj)

Arguments

Value

A list of SummarizedExperiment objects for each group in the input object. Each object in the list corresponds to one group of samples.

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)
example_obj_list <- split_groups(example_obj)

Description

This function takes a list of merged SummarizedExperiment objects, which are the output of the calc_feature_property() function, and summarizes the feature properties across all groups. It extracts the row data from each SummarizedExperiment object, combines them into a single data frame, and returns this summary data frame. Each row in the resulting data frame represents a feature, and the columns contain the properties of that feature for each group, including the feature name, group name, proportion of NA values for that feature in that group, randomness p-value, and maximum fold change over time.

Usage

summarise_feature_property(se_obj_merged_list)

Arguments

Value

A data frame summarizing the feature properties across all groups, with each row representing a feature and columns containing the properties including the feature name, group name, and the proportion of NA values for that feature in that group, randomness p-value, and maximum fold change over time.

Examples

data(example_obj)
example_obj <- normalise_to_start(example_obj)
example_obj_list <- split_groups(example_obj)
example_obj_merged_list <- merge_replicates(example_obj_list)

example_obj_merged_list <-
    calc_feature_property(example_obj_merged_list, threshold = 0)
property_random_fc <- summarise_feature_property(example_obj_merged_list)

Description

Computes per-module metrics for a WGCNA network: module size, proportion of features assigned, and module membership (kME). Unassigned module is excluded.

Usage

summarise_module_metrics(net)

Arguments

Details

MeanKME and MeanKME2 are the mean and mean squared module membership (kME = cor(feature, module eigengene)). MeanKME2 = mean(kME^2) is the average proportion of per-feature variance explained by the module eigengene.

Value

A data frame with columns: Module, Size, Proportion, MeanKME, MeanKME2, MedianKME, SDKME, MinKME, MaxKME.

Examples

data(example_net)
summarise_module_metrics(example_net)

Description

Summarise module patterns by integrating WGCNA output with Trendy results

Usage

summarise_module_pattern(module, trendy_summary)

Arguments

Value

A list of data frames, each data frame shows the pattern counts in a module

Examples

data("example_res_list")
trendy_summary <- summarise_Trendy(example_res_list)

data(example_net)
example_module <- WGCNA_module(example_net)
summarise_module_pattern(example_module, trendy_summary)

Description

Summarise Trendy results into data frame

Usage

summarise_Trendy(res_list, ...)

Arguments

Value

A data frame of summary results of Trendy, including breakpoints, segment slopes and p-values for each fitted feature in each group. A "Pattern" column is added to show the combination of segment trends (up, down, stable) for each feature.

Examples

data("example_res_list")
trendy_summary <- summarise_Trendy(example_res_list)

Description

A minimal theme used by all TiDEomics plotting functions, built on theme_minimal(). Exported so users can apply it to their own plots for consistent styling.

Usage

theme_custom(base_size = 8, panel_border = FALSE, legend_position = "right")

Arguments

Value

A ggplot2 theme object.

Examples

library(ggplot2)
ggplot(mtcars, aes(wt, mpg)) +
    geom_point() +
    theme_custom(base_size = 10)

Description

A subset of GSE263759 data set published in Integrated time-series analysis and high-content CRISPR screening delineate the dynamics of macrophage immune regulation

Usage

data(tutorial_data)

Format

A data.frame with 500 rows and 41 variables:

Feature

Feature ID, e.g. gene symbols

Sample1, Sample2, ...

log2(CPM + 1) normalised expression values

Details

Code for preparing the data is available in inst/script/tutorial_input.R

• Ensembl IDs were mapped to symbols, genes with all zero counts were excluded.

• Use time 0 untreated samples for other groups' time 0.

• Include "untreated", "IFNbeta", "IFNgamma", "LPS" groups.

• Sample 500 random genes.

• Normalised to log2(CPM + 1).

Source

GSE263759

Description

A subset of GSE263759 data set published in Integrated time-series analysis and high-content CRISPR screening delineate the dynamics of macrophage immune regulation

Usage

data(tutorial_sample_info)

Format

A data.frame with 40 rows and 5 variables:

Sample

Sample ID

Group

Experimental group (untreated, different treatments)

Time

Time point

Replicate

Replicate ID for each group and time point

Batch

Batch information

Details

Code for preparing the data is available in inst/script/tutorial_input.R

• Ensembl IDs were mapped to symbols, genes with all zero counts were excluded.

• Use time 0 untreated samples for other groups' time 0.

• Include "untreated", "IFNbeta", "IFNgamma", "LPS" groups.

• Sample 500 random genes.

Source

GSE263759

Description

Converts the module assignments from run_WGCNA() output into two-column data.frame with Feature and Module columns, expected by plot_modules_v(), plot_modules_h(), summarise_module_pattern(), and all enrichment functions.

Usage

WGCNA_module(net, exclude_grey = FALSE)

Arguments

Value

A data.frame with columns Feature (character) and Module (factor ordered by decreasing module size). When exclude_grey = TRUE, grey/unassigned features are excluded.

Examples

data(example_net)
module <- WGCNA_module(example_net)