Package 'pmartR'

Description

Test if a file is an edata file

Usage

.is_edata(edata)

Arguments

Value

A boolean where TRUE means the file is an acceptable edata file.

Description

Selects biomolecules for normalization via choosing all biomolecules currently in the data

Usage

all_subset(e_data, edata_id)

Arguments

Details

This function returns the subset of all biomolecules. These will be used for normalization.

Value

Character vector containing all biomolecules.

Author(s)

Kelly Stratton

Description

The method identifies biomolecules to be filtered specifically according data requirements for running an ANOVA.

Usage

anova_filter(nonmiss_per_group, min_nonmiss_anova, comparisons = NULL)

Arguments

Details

This function filters biomolecules that do not have at least min.nonmiss.allowed values per group, where groups are from group_designation.

Value

filter.peps a character vector of the biomolecules to be filtered out prior to ANOVA or IMD-ANOVA

Author(s)

Kelly Stratton

See Also

nonmissing_per_group

Description

This is the ANOVA part of the IMD-ANOVA test proposed in Webb-Robertson et al. (2010).

Usage

anova_test(
  omicsData,
  groupData,
  comparisons,
  pval_adjust_multcomp,
  pval_adjust_fdr,
  pval_thresh,
  covariates,
  paired,
  equal_var,
  parallel
)

Arguments

Details

The order in which different scenarios are handeled:

1. If the data are paired, then the pairing is accounted for first then each of the next steps is carried out on the new variable that is the difference in the paired individuals.<br>

2. If covariates are provided, their effect is removed before testing for group differences though mathematically covariates and grouping effects are accounted for simultaneously

3. ANOVA is executed to assess the effect of each main effects, results in a vector of group means for each biomolecule and variance estimate

4. Group comparisons defined by 'comaprison' argument are implemented use parameter vector and variance estimates in ANOVA step

Value

a list of 'data.frame's

Author(s)

Bryan Stanfill, Daniel Claborne

References

Webb-Robertson, Bobbie-Jo M., et al. "Combined statistical analyses of peptide intensities and peptide occurrences improves identification of significant peptides from MS-based proteomics data." Journal of proteome research 9.11 (2010): 5748-5756.

Description

This function takes a filter object of class 'cvFilt', 'rmdFilt', 'moleculeFilt', 'proteomicsFilt', 'imdanovaFilt', 'RNAFilt', 'totalCountFilt', or 'customFilt' and applies the filter to a dataset of pepData, proData, lipidData, metabData, nmrData or seqData.

Usage

applyFilt(filter_object, omicsData, ...)

## S3 method for class 'moleculeFilt'
applyFilt(filter_object, omicsData, min_num = 2, ...)

## S3 method for class 'totalCountFilt'
applyFilt(filter_object, omicsData, min_count, ...)

## S3 method for class 'RNAFilt'
applyFilt(
  filter_object,
  omicsData,
  min_nonzero = NULL,
  size_library = NULL,
  ...
)

## S3 method for class 'cvFilt'
applyFilt(filter_object, omicsData, cv_threshold = 150, ...)

## S3 method for class 'rmdFilt'
applyFilt(
  filter_object,
  omicsData,
  pvalue_threshold = 1e-04,
  min_num_biomolecules = 50,
  ...
)

## S3 method for class 'proteomicsFilt'
applyFilt(
  filter_object,
  omicsData,
  min_num_peps = NULL,
  redundancy = FALSE,
  ...
)

## S3 method for class 'imdanovaFilt'
applyFilt(
  filter_object,
  omicsData,
  comparisons = NULL,
  min_nonmiss_anova = NULL,
  min_nonmiss_gtest = NULL,
  remove_singleton_groups = TRUE,
  ...
)

## S3 method for class 'customFilt'
applyFilt(filter_object, omicsData, ...)

Arguments

Details

Further arguments can be specified depending on the class of the filter_object being applied.

For a filter_object of type 'moleculeFilt':

For a filter_object of type 'cvFilt':

For a filter_object of type 'rmdFilt':

For a filter_object of type 'proteomicsFilt' either or both of the following can be applied:

For a filter_object of type 'imdanovaFilt':

For a filter_object of type 'totalCountFilt':

For a filter_object of type 'RNAFilt' either or both of the following can be applied:

There are no further arguments for a filter_object of type ' customFilt'.

Value

An object of the class pepData, proData, lipidData, metabData, nmrData, or seqData with specified cname_ids, edata_cnames, and emeta_cnames filtered out of the appropriate datasets.

Author(s)

Lisa Bramer, Kelly Stratton

See Also

molecule_filter

imdanova_filter

rmd_filter

cv_filter

proteomics_filter

custom_filter

RNA_filter

total_count_filter

Examples

library(pmartRdata)
to_filter <- molecule_filter(omicsData = pep_object)
summary(to_filter, min_num = 2)
pep_object2 <- applyFilt(
  filter_object = to_filter,
  omicsData = pep_object, min_num = 2
)
summary(pep_object2) # number of Peptides is as expected based on summary of 
                     # the filter object that was applied
pep_object2 <- group_designation(omicsData = pep_object2,
                                 main_effects = "Phenotype")
to_filter2 <- imdanova_filter(omicsData = pep_object2)
pep_object3 <- applyFilt(
  filter_object = to_filter2,
  omicsData = pep_object2,
  min_nonmiss_anova = 3
)

Description

Converts several data frames of isobaric peptide data to an object of the class 'isobaricpepData'. Objects of the class 'isobaricpepData' are lists with two obligatory components, e_data and f_data. An optional list component, e_meta, is used if analysis or visualization at other levels (e.g. protein) is also desired.

Usage

as.isobaricpepData(
  e_data,
  f_data,
  e_meta = NULL,
  edata_cname,
  fdata_cname,
  emeta_cname = NULL,
  techrep_cname = NULL,
  ...
)

Arguments

Details

The class 'isobaricpepData' is meant to deal with labeled peptide data generated on instruments (e.g. TMT, iTRAQ) where a reference pool sample will be utilized for normalization.

If your data has already undergone normalization to the reference pool, you should specify isobaric_norm = T.

Objects of class 'isobaricpepData' contain some attributes that are referenced by downstream functions. These attributes can be changed from their default value by manual specification. A list of these attributes as well as their default values are as follows:

Computed values included in the data_info attribute are as follows:

Value

Object of class isobaricpepData and pepData.

Author(s)

Lisa Bramer

See Also

as.pepData

normalize_isobaric

Examples

library(pmartRdata)
mypep <- as.isobaricpepData(
  e_data = isobaric_edata,
  e_meta = isobaric_emeta,
  f_data = isobaric_fdata,
  edata_cname = "Peptide",
  fdata_cname = "SampleID",
  emeta_cname = "Protein"
)

Description

Converts several data frames of lipid data to an object of the class 'lipidData'. Objects of the class 'lipidData' are lists with two obligatory components, e_data and f_data. An optional list component, e_meta, is used if analysis or visualization at other levels (e.g. lipid class) is also desired.

Usage

as.lipidData(
  e_data,
  f_data,
  e_meta = NULL,
  edata_cname,
  fdata_cname,
  emeta_cname = NULL,
  techrep_cname = NULL,
  ...
)

Arguments

Details

Objects of class 'lipidData' contain some attributes that are referenced by downstream functions. These attributes can be changed from their default value by manual specification. A list of these attributes as well as their default values are as follows:

Computed values included in the data_info attribute are as follows:

Value

Object of class lipidData

Author(s)

Lisa Bramer, Kelly Stratton

See Also

as.metabData

Examples

library(pmartRdata)
mylipid <- as.lipidData(
  e_data = lipid_neg_edata,
  f_data = lipid_neg_fdata,
  edata_cname = "Lipid",
  fdata_cname = "SampleID"
)

Description

Converts several data frames of metabolomic data to an object of the class 'metabData'. Objects of the class 'metabData' are lists with two obligatory components, e_data and f_data. An optional list component, e_meta, is used if analysis or visualization at other levels (e.g. metabolite identification) is also desired.

Usage

as.metabData(
  e_data,
  f_data,
  e_meta = NULL,
  edata_cname,
  fdata_cname,
  emeta_cname = NULL,
  techrep_cname = NULL,
  ...
)

Arguments

Details

Objects of class 'metabData' contain some attributes that are referenced by downstream functions. These attributes can be changed from their default value by manual specification. A list of these attributes as well as their default values are as follows:

Computed values included in the data_info attribute are as follows:

Value

Object of class metabData

Author(s)

Lisa Bramer, Kelly Stratton

See Also

as.lipidData

as.nmrData

Examples

library(pmartRdata)
mymetabData <- as.metabData(
  e_data = metab_edata,
  f_data = metab_fdata,
  edata_cname = "Metabolite",
  fdata_cname = "SampleID"
)

Description

Create a 'multiData' object from multiple omicsData objects

Usage

as.multiData(
  ...,
  f_meta = NULL,
  sample_intersect = FALSE,
  match_samples = TRUE,
  keep_sample_info = FALSE,
  auto_fmeta = FALSE
)

Arguments

Details

Object limits: Currently, as.multiData accepts at most one object from each of classes 'pepData/proData', 'metabData', 'nmrData', and at most two objects of class 'lipidData'.

sample_intersect will auto-align samples that occur in all datasets. Specifically, it creates a vector of all samples that are common across all datasets, and simply creates an f_meta by copying this vector for each dataset and column-binding them.

Value

Object of class 'multiData' containing the omicsData objects, and the sample alignment information f_meta.

See Also

combine_lipidData if you want to combine lipidData objects before providing them to as.multiData.

Examples

library(pmartRdata)

# Combine metabolomics and protein object into multidata, both must be log2
# and normalized.
mymetab <- edata_transform(omicsData = metab_object, data_scale = "log2")
mymetab <- normalize_global(omicsData = mymetab, subset_fn = "all",
                            norm_fn = "median", apply_norm = TRUE)

mypro <- pro_object

# Combine without specifically supplying f_meta, either directly, or as one
# of the f_datas in any object.
mymultidata <- as.multiData(mymetab, mypro, auto_fmeta = TRUE, sample_intersect = TRUE)

# Manually supply an f_meta
f_meta <- data.frame(
  "Proteins" = mypro$f_data$SampleID[match(mymetab$f_data$SampleID, mypro$f_data$SampleID)],
  "Metabolites" = mymetab$f_data$SampleID,
  "Condition" = mymetab$f_data$Phenotype[match(mymetab$f_data$SampleID, mypro$f_data$SampleID)]
)

mymultidata <- as.multiData(mymetab, mypro, f_meta = f_meta)
# remove samples that are not common across all data.
mymultidata <- as.multiData(mymetab, mypro, f_meta = f_meta, sample_intersect = TRUE)

Description

Converts several data frames of NMR-generated metabolomic data to an object of the class 'nmrData'. Objects of the class 'nmrData' are lists with two obligatory components, e_data and f_data. An optional list component, e_meta, is used if analysis or visualization at other levels (e.g. metabolite identification) is also desired.

Usage

as.nmrData(
  e_data,
  f_data,
  e_meta = NULL,
  edata_cname,
  fdata_cname,
  emeta_cname = NULL,
  techrep_cname = NULL,
  ...
)

Arguments

Details

Objects of class 'nmrData' contain some attributes that are referenced by downstream functions. These attributes can be changed from their default value by manual specification. A list of these attributes as well as their default values are as follows:

Computed values included in the data_info attribute are as follows:

Value

Object of class nmrData

Author(s)

Lisa Bramer, Kelly Stratton

See Also

as.metabData

normalize_nmr

Examples

library(pmartRdata)
mynmrData <- as.nmrData(
  e_data = nmr_identified_edata,
  f_data = nmr_identified_fdata,
  edata_cname = "Metabolite",
  fdata_cname = "SampleID",
  data_type = "identified"
)

Description

Converts several data frames of (unlabeled or global, as opposed to labeled) peptide data to an object of the class 'pepData'. Objects of the class 'pepData' are lists with two obligatory components, e_data and f_data. An optional list component, e_meta, is used if analysis or visualization at other levels (e.g. protein) is also desired.

Usage

as.pepData(
  e_data,
  f_data,
  e_meta = NULL,
  edata_cname,
  fdata_cname,
  emeta_cname = NULL,
  techrep_cname = NULL,
  ...
)

Arguments

Details

Objects of class 'pepData' contain some attributes that are referenced by downstream functions. These attributes can be changed from their default value by manual specification. A list of these attributes as well as their default values are as follows:

Computed values included in the data_info attribute are as follows:

Value

Object of class pepData

Author(s)

Kelly Stratton, Lisa Bramer

See Also

as.proData

as.isobaricpepData

Examples

library(pmartRdata)
mypepData <- as.pepData(
  e_data = pep_edata,
  e_meta = pep_emeta,
  f_data = pep_fdata,
  edata_cname = "Peptide",
  fdata_cname = "SampleID",
  emeta_cname = "RazorProtein"
)

Description

Converts several data frames of protein data to an object of the class 'proData'. Objects of the class 'proData' are lists with two obligatory components, e_data and f_data. An optional list component, e_meta, is used if analysis or visualization at other levels (e.g. gene) is also desired.

Usage

as.proData(
  e_data,
  f_data,
  e_meta = NULL,
  edata_cname,
  fdata_cname,
  emeta_cname = NULL,
  techrep_cname = NULL,
  ...
)

Arguments

Details

Objects of class 'proData' contain some attributes that are referenced by downstream functions. These attributes can be changed from their default value by manual specification. A list of these attributes as well as their default values are as follows:

Computed values included in the data_info attribute are as follows:

Value

Object of class proData

Author(s)

Kelly Stratton, Lisa Bramer

See Also

as.pepData

as.isobaricpepData

Examples

library(pmartRdata)
myproData <- as.proData(
  e_data = pro_edata,
  f_data = pro_fdata,
  edata_cname = "RazorProtein",
  fdata_cname = "SampleID",
  is_normalized = TRUE
)

Description

Converts several data frames of RNA-seq transcript data to an object of the class 'seqData'. Objects of the class 'seqData' are lists with two obligatory components, e_data and f_data. An optional list component, e_meta, is used if analysis or visualization at other levels (e.g. gene, protein, pathway) is also desired.

Usage

as.seqData(
  e_data,
  f_data,
  e_meta = NULL,
  edata_cname,
  fdata_cname,
  emeta_cname = NULL,
  techrep_cname = NULL,
  ...
)

Arguments

Details

Objects of class 'seqData' contain some attributes that are referenced by downstream functions. These attributes can be changed from their default value by manual specification. A list of these attributes as well as their default values are as follows:

Computed values included in the data_info attribute are as follows:

Value

Object of class seqData

Author(s)

Rachel Richardson, Kelly Stratton, Lisa Bramer

See Also

as.proData

as.pepData

as.lipidData

as.metabData

as.nmrData

Examples

library(pmartRdata)
myseq <- as.seqData(
  e_data = rnaseq_edata,
  e_meta = rnaseq_emeta,
  f_data = rnaseq_fdata,
  edata_cname = "Transcript",
  fdata_cname = "SampleName",
  emeta_cname = "Transcript"
)

Description

Either an omicData and/or a statRes object are accepted. omicData must be transformed and normalized, unless the data is isobaric protein or NMR data. If group_designation() has been run on the omicData object to add "main_effects", the resulting plots will include groups. The main effects group_designation and e_meta columns are merged to the e_data in long format to create the trelliData.omics dataframe, and e_meta is merged to statRes in long format to create trelliData.stat dataframe.

Usage

as.trelliData(omicsData = NULL, statRes = NULL)

Arguments

Value

An object of class 'trelliData' containing the raw data and optionally, statRes. To be passed to trelliscope building functions.

Author(s)

David Degnan, Lisa Bramer

Examples

library(pmartRdata)

###########################
## MS/NMR OMICS EXAMPLES ##
###########################

# Transform the data
omicsData <- edata_transform(omicsData = pep_object, data_scale = "log2")

# Group the data by condition
omicsData <- group_designation(omicsData = omicsData, main_effects = c("Phenotype"))

# Apply the IMD ANOVA filter
imdanova_Filt <- imdanova_filter(omicsData = omicsData)
omicsData <- applyFilt(filter_object = imdanova_Filt, omicsData = omicsData,
                       min_nonmiss_anova = 2)

# Normalize my pepData
omicsData <- normalize_global(omicsData, "subset_fn" = "all", "norm_fn" = "median",
                             "apply_norm" = TRUE, "backtransform" = TRUE)

# Implement the IMD ANOVA method and compute all pairwise comparisons 
# (i.e. leave the `comparisons` argument NULL)
statRes <- imd_anova(omicsData = omicsData, test_method = 'combined')

# Generate the trelliData object
trelliData2 <- as.trelliData(omicsData = omicsData)
trelliData3 <- as.trelliData(statRes = statRes)
trelliData4 <- as.trelliData(omicsData = omicsData, statRes = statRes)

######################
## RNA-SEQ EXAMPLES ##  
######################

# Group data by condition
omicsData_seq <- group_designation(omicsData = rnaseq_object, main_effects = c("Virus"))

# Filter low transcript counts
omicsData_seq <- applyFilt(filter_object = total_count_filter(omicsData = omicsData_seq), 
 omicsData = omicsData_seq, min_count = 15)

# Select a normalization and statistics method (options are 'edgeR', 'DESeq2', and 'voom').
# See ?difexp_seq for more details
statRes_seq <- diffexp_seq(omicsData = omicsData_seq, method = "voom")

# Generate the trelliData object
trelliData_seq2 <- as.trelliData(omicsData = omicsData_seq)
trelliData_seq3 <- as.trelliData(statRes = statRes_seq)
trelliData_seq4 <- as.trelliData(omicsData = omicsData_seq, statRes = statRes_seq)

Description

The only acceptable input file type is a single edata file. Transformation and normalization must be specified. Isobaric protein or NMR data does not need to be normalized.

Usage

as.trelliData.edata(
  e_data,
  edata_cname,
  omics_type,
  data_scale_original = "abundance",
  data_scale = "log2",
  normalization_fun = "global",
  normalization_params = list(subset_fn = "all", norm_fn = "median", apply_norm = TRUE,
    backtransform = TRUE),
  is_normalized = FALSE,
  force_normalization = FALSE
)

Arguments

Value

An object of class 'trelliData' containing the raw data. To be passed to trelliscope building functions.

Author(s)

David Degnan, Daniel Claborne, Lisa Bramer

Examples

library(pmartRdata)

###########################
## MS/NMR OMICS EXAMPLES ##
###########################

# Simple MS/NMR Example 
trelliData1 <- as.trelliData.edata(e_data = pep_edata,
                                   edata_cname = "Peptide",
                                   omics_type = "pepData")
 
######################
## RNA-SEQ EXAMPLES ##  
######################
                                    
# RNA-seq Example
trelliData_seq1 <- as.trelliData.edata(e_data = rnaseq_edata, 
                                      edata_cname = "Transcript",
                                      omics_type = "seqData")

Description

Applies BP-Quant to a pepData object

Usage

bpquant(statRes, pepData, pi_not = 0.9, max_proteoforms = 5, parallel = TRUE)

Arguments

Details

The result of this function can be used as one the isoformRes input argument to protein_quant. The bpquant function itself operates as follows: The statRes object contains the signatures data frame, the pepData object is used for its e_meta data frame. Next the signatures data frame and e_meta are merged by their edata_cname (e.g. peptide identifier) columns, this new data frame called protein_sig_data will be input to bpquant_mod in a “foreach” statement. “Foreach” will subset protein_sig_data for each unique protein and apply bpquant_mod to each subset and store the results.

Value

a list of data frames, one for each unique protein. The data frames have three columns, a protein identifier, a peptide identifier, and a "ProteoformID". The class of this list is 'isoformRes'.

Examples

library(pmartRdata)

mypepData <- group_designation(
  omicsData = pep_object,
  main_effects = c("Phenotype")
)
mypepData = edata_transform(omicsData = mypepData, data_scale = "log2")

imdanova_Filt <- imdanova_filter(omicsData = mypepData)
mypepData <- applyFilt(
  filter_object = imdanova_Filt,
  omicsData = mypepData,
  min_nonmiss_anova = 2
)

imd_anova_res <- imd_anova(
  omicsData = mypepData,
  test_method = 'combined',
  pval_adjust_a_multcomp = 'bon',
  pval_adjust_g_multcomp = 'bon'
)

result = bpquant(statRes = imd_anova_res, pepData = mypepData)

Description

The function is written to take input from one protein at a time and requires three inputs: protein_sig, pi_not and max_proteforms

Usage

bpquant_mod(protein_sig, pi_not, max_proteoforms)

Arguments

Details

Value

a list of five items: num_proteoforms, unique_sigs, proteoform_configs, post_prob and peptide_idx

Description

Combines two omicsData objects with identical sample information.

Usage

combine_lipidData(
  obj_1,
  obj_2,
  retain_groups = FALSE,
  retain_filters = FALSE,
  drop_duplicate_emeta = TRUE,
  ...
)

Arguments

Details

General requirements:

* sample names: These must be identical for both objects (column names of e_data, and sample identifiers in f_data) * data attributes: Objects must be on the same scale and both be either normalized or unnormalized * group designation: Objects must have the same grouping structure if retain_groups = T

Value

An object of the same type as the two input objects, with their combined data.

Examples

library(pmartRdata)

obj_1 <- lipid_neg_object
obj_2 <- lipid_pos_object

# de-duplicate any duplicate edata identifiers
all(obj_2$e_data[, get_edata_cname(obj_2)] == obj_2$e_meta[, get_edata_cname(obj_2)])
obj_2$e_data[, get_edata_cname(obj_2)] <- paste0("obj_2_", obj_2$e_data[, get_edata_cname(obj_2)])
obj_2$e_meta[, get_edata_cname(obj_2)] <- obj_2$e_data[, get_edata_cname(obj_2)]

combine_object <- combine_lipidData(obj_1 = obj_1, obj_2 = obj_2)

# preprocess and group the data and keep filters/grouping structure

obj_1 <- edata_transform(omicsData = obj_1, data_scale = "log2")
obj_1 <- normalize_global(omicsData = obj_1, subset_fn = "all",
                          norm_fn = "median", apply_norm = TRUE)
obj_2 <- edata_transform(omicsData = obj_2, data_scale = "log2")
obj_2 <- normalize_global(omicsData = obj_2, subset_fn = "all",
                          norm_fn = "median", apply_norm = TRUE)

obj_1 <- group_designation(omicsData = obj_1, main_effects = "Virus")
obj_2 <- group_designation(omicsData = obj_2, main_effects = "Virus")

obj_1 <- applyFilt(filter_object = molecule_filter(omicsData = obj_1),
                   omicsData = obj_1, min_num = 2)
obj_2 <- applyFilt(filter_object = cv_filter(omicsData = obj_2), obj_2, cv_thresh = 60)

combine_object_later <- combine_lipidData(
                          obj_1 = obj_1,
                          obj_2 = obj_2,
                          retain_groups = TRUE,
                          retain_filters = TRUE
                        )

Description

For each biomolecule, this function aggregates the technical replicates of the biological samples using a specified aggregation method

Usage

combine_techreps(omicsData, combine_fn = NULL, bio_sample_names = NULL)

Arguments

Details

Loss of information after aggregation

Value

An object with the same class as omicsData that has been aggregated to the biological sample level

Author(s)

Daniel Claborne

Examples

library(pmartRdata)

pep_object_averaged <- combine_techreps(omicsData = pep_techrep_object)

Description

Selects biomolecules that have complete rows in e_data, equivalent to 'ppp' with proportion = 1.

Usage

complete_mols(e_data, edata_id)

Arguments

Value

Character vector containing the biomolecules with no missing values across all samples.

Description

This function returns an object of class corRes (correlation Result)

Usage

cor_result(omicsData)

Arguments

Details

The pairwise correlations between samples are calculated based on biomolecules that are observed in both samples. For seqData objects, Spearman correlation is used. For all other data types, Pearson correlation is used and data must be log transformed. See cor for further details.

Value

An $n \times n$ matrix of class corRes giving the correlation between samples.

Author(s)

Kelly Stratton, Lisa Bramer

See Also

edata_transform

Examples

library(pmartRdata)

mymetab <- edata_transform(omicsData = metab_object, data_scale = "log2")
my_correlation <- cor_result(omicsData = mymetab)


myseq_correlation <- cor_result(omicsData = rnaseq_object)

Description

The method creates a data frame containing the comparisons to be made when performing differential statistics.

Usage

create_comparisonDF(comp_type, omicsData, control_group = NULL)

Arguments

Details

This function takes in the omicsData and type of comparison, and returns a data frame where each row corresponds to a comparison of interest.

Value

data frame with columns for Test and Control. Each row corresponds to a comparison of interest.

Author(s)

Kelly Stratton

See Also

group_designation

Examples

library(pmartRdata)
mymetab <- group_designation(omicsData = metab_object, main_effects = "Phenotype")
create_comparisonDF(comp_type = "pairwise", omicsData = mymetab)

create_comparisonDF(comp_type = "control", omicsData = mymetab, control_group = "Phenotype1")

Description

This function creates a customFilt S3 object based on user-specified items to filter out of the dataset

Usage

custom_filter(
  omicsData,
  e_data_remove = NULL,
  f_data_remove = NULL,
  e_meta_remove = NULL,
  e_data_keep = NULL,
  f_data_keep = NULL,
  e_meta_keep = NULL
)

Arguments

Value

An S3 object of class 'customFilt', which is a list with 3 elements for e_data, f_data, and e_meta, specifying which entries should be either kept or removed

Author(s)

Kelly Stratton

Examples

library(pmartRdata)
to_filter <- custom_filter(omicsData = metab_object, 
                           e_data_remove = "fumaric acid",
                           f_data_remove = "Sample_1_Phenotype2_B")
summary(to_filter)

to_filter2 <- custom_filter(omicsData = metab_object, 
                            f_data_keep = metab_object$f_data$SampleID[1:10])
summary(to_filter2)

Description

This helper function creates custom sample names for plot data object functions

Usage

custom_sampnames(
  omicsData,
  firstn = NULL,
  from = NULL,
  to = NULL,
  delim = NULL,
  components = NULL,
  pattern = NULL,
  ...
)

Arguments

Details

This function can be used to create custom (and shorter) sample names to be used when plotting so that axis labels are not so long that they interfere with the graph itself. To use the custom sample names when plotting, specify the optional argument 'use_VizSampNames = TRUE'.

Value

Object of same class as omicsData, with added column in f_data named 'VizSampNames'.

Examples

library(pmartRdata)

mypep <- edata_transform(omicsData = pep_object, data_scale = "log2")
plot(mypep)

# specify new names using firstn argument
results <- custom_sampnames(omicsData = mypep, firstn = 9)
plot(results, use_VizSampNames = TRUE)

# specify new names using from and to arguments
results <- custom_sampnames(omicsData = mypep, from = 1, to = 9)
plot(results, use_VizSampNames = TRUE)

# specify new names using delim and components arguments
results <- custom_sampnames(omicsData = mypep, delim = "_", components = c(1, 2))
plot(results, use_VizSampNames = TRUE)

## specify new names using pattern arguments (regex)

# match everything after "Sample_"
pattern1 <- "[0-9]+_[0-9A-Za-z]+_[A-Z]"

results <- custom_sampnames(omicsData = mypep, pattern = pattern1)
plot(results, use_VizSampNames = TRUE)

# match "Sample_" and the number after it
pattern2 <- "^Sample_[0-9]+"

results <- custom_sampnames(omicsData = mypep, pattern = pattern2)
plot(results, use_VizSampNames = TRUE)

Description

A pooled CV is calculated for each biomolecule.

Usage

cv_filter(omicsData, use_groups = TRUE)

Arguments

Details

For each biomolecule, the CV of each group is calculated as the standard deviation divided by the mean, excluding missing values. A pooled CV estimate is then calculated based on the methods of Ahmed (1995). Any groups consisting of a single sample are excluded from the CV calculation, and thus, from the cv_filter result. If group_designation has not been run on the omicsData object, all samples are considered to belong to the same group.

Value

An S3 object of class 'cvFilt' giving the pooled CV for each biomolecule and additional information used for plotting a data.frame with a column giving the biomolecule name and a column giving the pooled CV value.

Author(s)

Lisa Bramer, Kelly Stratton

References

Ahmed, S.E. (1995). A pooling methodology for coefficient of variation. The Indian Journal of Statistics. 57: 57-75.

Examples

library(pmartRdata)
mypep <- group_designation(omicsData = pep_object, 
                           main_effects = "Phenotype")
to_filter <- cv_filter(omicsData = mypep, use_groups = TRUE)
summary(to_filter, cv_threshold = 30)

Description

For generating statistics for 'seqData' objects

Usage

DESeq2_wrapper(
  omicsData,
  test = "Wald",
  p_adjust = "BH",
  comparisons = NULL,
  p_cutoff = 0.05,
  ...
)

Arguments

Details

Runs default DESeq workflow. Defaults to Wald test, no independent filtering, and running in parallel. Additional arguments can be passed for use in the function, refer to DESeq() and results() in DESeq2 package. Requires 'survival' package to run.

Flags (signatures) - Indicator of statistical significance for computed test. Zeros indicate no significance, while +/- 1 indicates direction of significance.

Value

statRes object

References

Love, M.I., Huber, W., Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2 Genome Biology 15(12):550 (2014)

Description

Performs statistical analysis for differential expression of seqData objects, using methods from one of: edgeR, DESeq2, or limma-voom

Usage

diffexp_seq(
  omicsData,
  method = "edgeR",
  p_adjust = "BH",
  comparisons = NULL,
  p_cutoff = 0.05,
  ...
)

Arguments

Details

Runs default differential expression workflows.

Flags (signatures) - Indicator of statistical significance. Zeroes indicate no significance, while +/- 1 indicates direction of significance.

Method "edgeR" - Runs default edgeR workflow with empirical Bayes quasi-likelihood F-tests. Additional arguments can be passed for use in the function. Refer to calcNormFactors() and glmQLFit() in edgeR package. Requires the 'edgeR' and 'limma' packages to run.

Method "DESeq2" - Runs default DESeq workflow. Defaults to Wald test, no independent filtering, and running in parallel. Additional arguments can be passed for use in the function. Refer to DESeq() and results() in DESeq2 package. Requires 'survival' package to run.

Method "voom" - Runs default limma-voom workflow using empirical Bayes moderated t-statistics. Additional arguments can be passed for use in the function. Refer to calcNormFactors() in edgeR package. Requires the 'edgeR' and 'limma' packages to run.

Value

object of class statRes

References

Robinson MD, McCarthy DJ, Smyth GK (2010). “edgeR: a Bioconductor package for differential expression analysis of digital gene expression data.” Bioinformatics, 26(1), 139-140. doi: 10.1093/bioinformatics/btp616.

Love, M.I., Huber, W., Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2 Genome Biology 15(12):550 (2014)

Ritchie, M.E., Phipson, B., Wu, D., Hu, Y., Law, C.W., Shi, W., and Smyth, G.K. (2015). limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Research 43(7), e47.

Examples

library(pmartRdata)
myseqData <- group_designation(omicsData = rnaseq_object, main_effects = "Virus")
edger_results <- diffexp_seq(omicsData = myseqData, method = "edgeR")
deseq_results <- diffexp_seq(omicsData = myseqData, method = "DESeq2")
voom_results <- diffexp_seq(omicsData = myseqData, method = "voom")

Description

For data types other than seqData, this function calculates principal components using projection pursuit estimation, which implements an expectation-maximization (EM) estimation algorithm when data is missing. For seqData counts, a generalized version of principal components analysis for non-normally distributed data is calculated under the assumption of a negative binomial distribution with global dispersion.

Usage

dim_reduction(omicsData, k = 2)

Arguments

Details

Any biomolecules seen in only one sample or with a variance less than 1E-6 across all samples are not included in the PCA calculations. This function leverages code from pca and glmpca .

Value

a data.frame with first k principal component scores, sample identifiers, and group membership for each sample (if group designation was previously run on the data). The object is of class dimRes (dimension reduction Result).

References

Redestig H, Stacklies W, Scholz M, Selbig J, & Walther D (2007). pcaMethods - a bioconductor package providing PCA methods for incomplete data. Bioinformatics. 23(9): 1164-7.

Townes FW, Hicks SC, Aryee MJ, Irizarry RA (2019). Feature selection and dimension reduction for single-cell RNA-seq based on a multinomial model. Genome Biol. 20, 1–16.

Huang H, Wang Y, Rudin C, Browne EP (2022). Towards a comprehensive evaluation of dimension reduction methods for transcriptomic data visualization. Communications Biology 5, 719.

Examples

library(pmartRdata)

mylipid <- edata_transform(omicsData = lipid_neg_object, data_scale = "log2")
mylipid <- group_designation(omicsData = mylipid, main_effects = "Virus")
pca_lipids <- dim_reduction(omicsData = mylipid)


myseq <- group_designation(omicsData = rnaseq_object, main_effects = "Virus")
pca_seq <- dim_reduction(omicsData = myseq)

Description

For generating statistics for 'seqData' objects

Usage

dispersion_est(
  omicsData,
  method,
  interactive = FALSE,
  x_lab = NULL,
  x_lab_size = 11,
  x_lab_angle = NULL,
  y_lab = NULL,
  y_lab_size = 11,
  title_lab = NULL,
  title_lab_size = 14,
  legend_lab = NULL,
  legend_position = "right",
  bw_theme = TRUE,
  palette = NULL,
  point_size = 0.2,
  custom_theme = NULL
)

Arguments

Details

DESeq2 option requires package "survival" to be available.

Value

plot result

References

Robinson MD, McCarthy DJ, Smyth GK (2010). “edgeR: a Bioconductor package for differential expression analysis of digital gene expression data.” Bioinformatics, 26(1), 139-140. doi: 10.1093/bioinformatics/btp616.

Love, M.I., Huber, W., Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2 Genome Biology 15(12):550 (2014)

Ritchie, M.E., Phipson, B., Wu, D., Hu, Y., Law, C.W., Shi, W., and Smyth, G.K. (2015). limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Research 43(7), e47.

Examples

library(pmartRdata)
myseqData <- group_designation(omicsData = rnaseq_object, main_effects = "Virus")
dispersion_est(omicsData = myseqData, method = "edgeR")
dispersion_est(omicsData = myseqData, method = "DESeq2")
dispersion_est(omicsData = myseqData, method = "voom")

Description

This function finds all values of x in the e_data element of omicsData and replaces them with y

Usage

edata_replace(omicsData, x, y, threshold = NULL)

Arguments

Details

This function is often used to replace any 0 values in peptide, protein, metabolite, or lipid data with NA's. For omicsData on the abundance scale, when the omicsData object is created, any 0's in e_data are automatically converted to NA's. For omicsData on the count scale (e.g. seqData objects), when the omicsData object is created, any NA's in e_data are automatically converted to 0's.

Value

data object of the same class as omicsData

Author(s)

Kelly Stratton

Examples

library(pmartRdata)
mymetab <- edata_replace(omicsData = metab_object, x = 0, y = NA)

Description

This function takes in an omicsData object and returns a summary of the e_data component. The six summarizing metrics include the mean, standard deviation, median, percent observed, minimum, and maximum.

Usage

edata_summary(omicsData, by = "sample", groupvar = NULL)

Arguments

Details

If groupvar is NULL and group_designation has not been applied to omicsData, then the metrics will be applied to each column of e_data (when by = 'sample) or to each row of e_data (when by = 'molecule'). When groupvar is provided, it must match a column name from f_data, this column of f_data is used to group e_data in order to apply the metrics.

Value

A list of six data frames, of class 'dataRes' (data Result), which are the results of applying the metrics (mean, standard deviation, median, percent observed, minimum and maximum) to omicsData$e_data.

Examples

library(pmartRdata)

mylipid <- edata_transform(omicsData = lipid_pos_object, data_scale = "log2")
mylipid <- group_designation(omicsData = mylipid, main_effects = "Virus")
result <- edata_summary(omicsData = mylipid, by = "sample", groupvar = NULL)

Description

This function applies a transformation to the e_data element of omicsData

Usage

edata_transform(omicsData, data_scale)

Arguments

Details

For all but seqData, this function is intended to be used before analysis of the data begins, and data are typically analyzed on a log scale. This function is not applicable to seqData objects, as any transformations needed e.g. to allow more meaningful visualization of seqData objects are performed within the pertinent functions.

Value

data object of the same class as omicsData

Author(s)

Kelly Stratton, Natalie Heller

Examples

library(pmartRdata)
mymetab <- edata_transform(omicsData = metab_object, data_scale = "log2")
attr(mymetab, "data_info")$data_scale

Description

For generating statistics for 'seqData' objects.

Usage

edgeR_wrapper(
  omicsData,
  p_adjust = "BH",
  comparisons = NULL,
  p_cutoff = 0.05,
  ...
)

Arguments

Details

Requires the 'edgeR' and 'limma' packages. Runs default edgeR workflow with empirical Bayes quasi-likelihood F-tests. Additional arguments can be passed for use in the function, refer to calcNormFactors() and glmQLFit() in edgeR package.

Flags (signatures) - Indicator of statistical significance for computed test. Zeroes indicate no significance, while +/- 1 indicates direction of significance.

Value

statRes object

References

Robinson MD, McCarthy DJ, Smyth GK (2010). “edgeR: a Bioconductor package for differential expression analysis of digital gene expression data.” Bioinformatics, 26(1), 139-140. doi: 10.1093/bioinformatics/btp616.

Description

Implements overall survival analysis or progression-free survival analysis, depending upon the datatypes supplied to surv_designation, and return the "survfit" object

Usage

fit_surv(omicsData)

Arguments

Value

if fitted survival analysis object is returned

Examples

## Not run: 
library(MSomicsSTAT)
library(OvarianPepdataBP)

# Basic analysis without covariates
attr(tcga_ovarian_pepdata_bp, "survDF") <- list(t_death = "survival_time",
                                                ind_death = "vital_status")
sfit <- fit_surv(tcga_ovarian_pepdata_bp)
plot(sfit)

# Add some covariate information
attr(tcga_ovarian_pepdata_bp, "survDF") <- list(
  t_death = "survival_time",
  ind_death = "vital_status",
  covariates = "g__initial_pathologic_diagnosis_method_g1"
)
sfit <- fit_surv(tcga_ovarian_pepdata_bp)
plot(sfit, col = c(1, 2))

## End(Not run)

Description

Retrieves the value in check.names attribute from an omicsData object. This function has been deprecated in favor of handling checking names externally, and will always return FALSE.

Usage

get_check_names(omicsData)

Arguments

Value

A logical value indicating if the syntax of the column names in a data frame should be checked. See data.frame for more details.

Description

This function returns comparisons from statRes or trelliData object

Usage

get_comparisons(compObj)

Arguments

Value

returns a data frame with comparisons and their indices

Examples

library(pmartRdata)

my_prodata = group_designation(
  omicsData = pro_object,
  main_effects = c("Phenotype")
)

imdanova_Filt = imdanova_filter(omicsData = my_prodata)

my_prodata = applyFilt(
  filter_object = imdanova_Filt,
  omicsData = my_prodata,
  min_nonmiss_anova = 2
)

imd_anova_res = imd_anova(
  omicsData = my_prodata,
  test_method = 'comb',
  pval_adjust_a_multcomp = 'bon',
  pval_adjust_g_multcomp = 'bon'
)

result = get_comparisons(imd_anova_res)

Description

This function returns data_class attribute from statRes or trelliData object, inherited from the omicsData used in imd_anova or as.trelliData

Usage

get_data_class(dcObj)

Arguments

Value

returns the data_class attribute from a 'statRes' or 'trelliData' object

Examples

library(pmartRdata)

my_prodata = group_designation(
  omicsData = pro_object,
  main_effects = c("Phenotype")
)

imdanova_Filt = imdanova_filter(omicsData = my_prodata)

my_prodata = applyFilt(
  filter_object = imdanova_Filt,
  omicsData = my_prodata,
  min_nonmiss_anova = 2
)

imd_anova_res = imd_anova(
  omicsData = my_prodata,
  test_method = 'comb',
  pval_adjust_a_multcomp = 'bon',
  pval_adjust_g_multcomp = 'bon'
)

result = get_data_class(imd_anova_res)

Description

Retrieves the values in the data_info attribute from an omicsData object.

Usage

get_data_info(omicsData)

Arguments

Value

A list containing seven elements:

• data_scale – A Character string indicating the scale of the data in e_data.

• norm_info – A list containing a single element indicating whether the data in e_data have been normalized.

• num_edata – The number of unique entries present in the edata_cname column in e_data.

• num_miss_obs – An integer. The number of missing observations in e_data.

• num_zero_obs – An integer. The number of zeros in e_data for seqData objects.

• prop_missing – A number between 0 and 1. The proportion of missing observations in e_data.

• num_samps – An integer indicating the number of samples or columns (excluding the identifier column edata_cname) in e_data.

• data_types – A character string describing the type of data in e_data.

Description

This function returns the norm_info element of the data_info attribute indicating whether the data have been normalized.

Usage

get_data_norm(omicsObject)

Arguments

Value

A logical value indicating whether the data have been normalized.

Description

This function returns current data scale which may be different from the original data scale (if edata_transform was used).

Usage

get_data_scale(omicsObject)

Arguments

Value

a character string describing data scale

Description

Retrieves the character string indicating the scale the data was originally on when read into R.

Usage

get_data_scale_orig(omicsObject)

Arguments

Value

A character string.

Description

This function returns the name of the column in e_data that contains the biomolecule IDs.

Usage

get_edata_cname(omicsObject)

Arguments

Value

a character string describing e_data cname

Description

This function returns the name of the column in e_meta that contains the mapping variable IDs.

Usage

get_emeta_cname(omicsObject)

Arguments

Value

a character string describing e_meta cname

Description

This function returns the name of the column in f_data that contains the names of the samples.

Usage

get_fdata_cname(omicsObject)

Arguments

Value

a character string describing f_data cname

Description

Retrieves the values in the filters attribute from an omicsData object.

Usage

get_filter_type(omicsData)

Arguments

Value

vector of filters used on omicsData

Description

Retrieves the values in the filters attribute from an omicsData object.

Usage

get_filters(omicsData)

Arguments

Value

A list containing filter class objects. Each element in this list corresponds to a filter applied to the data. The filters will be listed in the order they were applied. A filter object contains two elements:

• threshold – The threshold used to filter e_data. This value depends on the type of filter applied.

• filtered – A vector containing the identifiers from the edata_cname column that will be filtered.

• method – A character string indicating the type of method used to filter. This only applies when imdanova_filter is used.

Description

Retrieves the values in the group_DF attribute from an omicsData object.

Usage

get_group_DF(omicsData)

Arguments

Value

A data.frame with columns for sample ID and group. If two main effects are provided the original main effect levels for each sample are returned as the third and fourth columns of the data frame. Additionally, the covariates provided will be listed as attributes of this data frame.

Description

For generating group design formulas and correctly ordered group data for seqData statistical functions

Usage

get_group_formula(omicsData)

Arguments

Value

A list with two elements:

• grouping_info: A data.frame with the grouping information used in the statistical analysis

• formula_string: A character string with the formula used in the statistical analysis

Description

This function returns a table with number of samples per group

Usage

get_group_table(omicsObject)

Arguments

Value

a table containing number of samples per group

Description

Retrieves the values in the isobaric_info attribute from an omicsData object.

Usage

get_isobaric_info(omicsData)

Arguments

Value

A list containing the following six elements:

• exp_cname –

• channel_cname –

• refpool_channel –

• refpool_cname –

• refpool_notation –

• norm_info – A list containing a single logical element that indicates whether the data have been normalized to a reference pool.

Description

This function returns the norm_info element of the isobaric_info attribute which indicates if the data have been isobaric normalized.

Usage

get_isobaric_norm(omicsData)

Arguments

Value

A logical value indicating whether the data have been isobaric normalized.

Description

Compute the least squares means from a prediction grid and estimated coefficients

Usage

get_lsmeans(data, xmatrix, pred_grid, Betas, continuous_covar_inds = NULL)

Arguments

Value

A data frame of the least squares means

Author(s)

Daniel Claborne

Description

Retrieves the values in the meta_info attribute from an omicsData object.

Usage

get_meta_info(omicsData)

Arguments

Value

A list containing two elements:

• meta_data – Logical. Indicates if the e_meta data frame was provided.

• num_emeta – The number of unique entries present in the emeta_cname column in e_meta.

Description

Retrieves the values in the nmr_info attribute from an omicsData object.

Usage

get_nmr_info(omicsData)

Arguments

Value

A list containing the following three elements:

• metabolite_name –

• sample_property_cname –

• norm_info – A list containing two logical elements that indicate i) whether the data have been normalized to a spiked in metabolite or to a property taking sample-specific values and ii) whether the data have been back transformed so the values are on a similar scale to the raw values before normalization.

Description

This function returns the norm_info element of the nmr_info attribute which indicates if the data have been NMR normalized.

Usage

get_nmr_norm(omicsData)

Arguments

Value

A logical value indicating whether the data have been NMR normalized.

Description

Build the prediction grid to compute least squares means.

Usage

get_pred_grid(
  group_df,
  main_effect_names,
  covariate_names = NULL,
  fspec = as.formula("~.")
)

Arguments

Value

A matrix of the prediction grid

Author(s)

Daniel Claborne

Description

Gets the parameters for the highest ranked methods from spans.

Usage

get_spans_params(SPANSRes_obj, sort_by_nmols = FALSE)

Arguments

Value

A list of lists, where there are multiple sublists only if there were ties for the top SPANS score. Each sublist contains named elements for the subset and normalization methods, and the parameters used for the subset method.

Examples

library(pmartRdata)

# data must be log transformed and grouped
myobject <- edata_transform(omicsData = pep_object, data_scale = "log2")
myobject <- group_designation(omicsData = myobject, main_effects = "Phenotype")

spans_result <- spans_procedure(omicsData = myobject)

# a list of the parameters for any normalization procedure with the best SPANS score
best_params <- get_spans_params(spans_result)

# extract the arguments from the first list element
subset_fn = best_params[[1]]$subset_fn
norm_fn = best_params[[1]]$norm_fn
params = best_params[[1]]$params
if (is.null(params[[1]])) {
  params = NULL
}

# pass arguments to normalize global
norm_object <- normalize_global(omicsData = myobject, subset_fn = subset_fn,
                                norm_fn = norm_fn, params = params)

Description

Takes the results of anova_test() and returns group comparison p-values

Usage

group_comparison_anova(
  data,
  groupData,
  comparisons,
  Xfull,
  Xred,
  anova_results_full,
  beta_to_mu_full,
  beta_to_mu_red
)

Arguments

Value

A data.frame containing the p-values from the group comparisons.

Author(s)

Bryan Stanfill, Daniel Claborne

Description

Takes the results of imd_test() and returns group comparison p-values

Usage

group_comparison_imd(groupData, comparisons, observed, absent)

Arguments

Value

A data.frame containing the p-values from the group comparisons.

Author(s)

Bryan Stanfill

Description

The method assigns each sample to a group, for use in future analyses, based on the variable(s) specified as main effects.

Usage

group_designation(
  omicsData,
  main_effects = NULL,
  covariates = NULL,
  cov_type = NULL,
  pair_id = NULL,
  pair_group = NULL,
  pair_denom = NULL,
  batch_id = NULL
)

Arguments

Details

Groups are formed based on the levels of the main effect variables. One or two main effect variables are allowed. In the case of two main effect variables, groups are formed based on unique combinations of the levels of the two main effect variables. Any samples with level NA for a main effect variable will be removed from the data and will not be included in the final group designation results. Groups with a single sample are allowed, as is a single group.

Value

An object of the same class as the input omicsData object - the provided object with the samples filtered out, if any NAs were produced in designating groups. An attribute 'group_DF', a data.frame with columns for sample id and group, is added to the object. If two main effects are provided the original main effect levels for each sample are returned as the third and fourth columns of the data.frame. Additionally, the covariates provided will be listed as attributes of this data.frame.

Author(s)

Lisa Bramer, Kelly Stratton

Examples

library(pmartRdata)
mylipid <- group_designation(
  omicsData = lipid_pos_object,
  main_effects = "Virus"
)
attr(mylipid, "group_DF")

Description

The method identifies peptides, proteins, lipids, or metabolites to be filtered specifically according to the G-test.

Usage

gtest_filter(nonmiss_per_group, min_nonmiss_gtest, comparisons = NULL)

Arguments

Details

Two methods are available for determining the peptides to be filtered. The naive approach is based on min.nonmiss.allowed, and looks for peptides that do not have at least min.nonmiss.allowed values per group. The other approach also looks for peptides that do not have at least a minimum number of values per group, but this minimum number is determined using the G-test and a p-value threshold supplied by the user. The G-test is a test of independence, used here to test the null hypothesis of independence between the number of missing values across groups.

Value

filter.peps a character vector of the peptides to be filtered out prior to the G-test or IMD-ANOVA

Author(s)

Kelly Stratton

Description

This is the IMD-ANOVA test defined in Webb-Robertson et al. (2010).

Usage

imd_anova(
  omicsData,
  comparisons = NULL,
  test_method,
  pval_adjust_a_multcomp = "none",
  pval_adjust_g_multcomp = "none",
  pval_adjust_a_fdr = "none",
  pval_adjust_g_fdr = "none",
  pval_thresh = 0.05,
  equal_var = TRUE,
  parallel = TRUE
)

Arguments

Value

An object of class 'statRes', which is a data frame containing columns (when relevant based on the test(s) performed) for: e_data cname, group counts, group means, ANOVA p-values, IMD p-values, fold change estimates on the same scale as the data (e.g. log2, log10, etc.), and fold change significance flags (0 = not significant; +1 = significant and positive fold change (ANOVA) or more observations in test group relative to reference group (IMD); -1 = significant and negative fold change (ANOVA) or fewer observations in test group relative to reference group (IMD))

Author(s)

Bryan Stanfill, Kelly Stratton

References

Webb-Robertson, Bobbie-Jo M., et al. "Combined statistical analyses of peptide intensities and peptide occurrences improves identification of significant peptides from MS-based proteomics data." Journal of proteome research 9.11 (2010): 5748-5756.

Examples

library(pmartRdata)
# Transform the data
mymetab <- edata_transform(omicsData = metab_object, data_scale = "log2")

# Group the data by condition
mymetab <- group_designation(omicsData = mymetab, main_effects = c("Phenotype"))

# Apply the IMD ANOVA filter
imdanova_Filt <- imdanova_filter(omicsData = mymetab)
mymetab <- applyFilt(filter_object = imdanova_Filt, omicsData = mymetab, min_nonmiss_anova = 2)

# Implement IMD ANOVA and compute all pairwise comparisons 
# (i.e. leave the comparisons argument NULL), with FDR adjustment
anova_res <- imd_anova(omicsData = mymetab, test_method = "anova",
                       pval_adjust_a_multcomp = "Holm", pval_adjust_a_fdr = "BY")
imd_res <- imd_anova(omicsData = mymetab, test_method = "gtest",
                     pval_adjust_g_multcomp = "bon", pval_adjust_g_fdr = "BY")
imd_anova_res <- imd_anova(omicsData = mymetab, test_method = "combined",
                           pval_adjust_a_fdr = "BY", pval_adjust_g_fdr = "BY")
imd_anova_res <- imd_anova(omicsData = mymetab, test_method = "combined",
                           pval_adjust_a_multcomp = "bon", pval_adjust_g_multcomp = "bon",
                           pval_adjust_a_fdr = "BY", pval_adjust_g_fdr = "BY")

# Two main effects and a covariate
mymetab <- group_designation(omicsData = mymetab, main_effects = c("Phenotype", "SecondPhenotype"),
                             covariates = "Characteristic")
imd_anova_res <- imd_anova(omicsData = mymetab, test_method = 'comb')

# Same but with custom comparisons
comp_df <- data.frame(Control = c("Phenotype1", "A"), Test = c("Phenotype2", "B"))
custom_comps_res <- imd_anova(omicsData = mymetab, comparisons = comp_df, test_method = "combined")

Description

Tests the null hypothesis that the number of missing observations is independent of the groups. A g-test is used to test this null hypothese against the alternative that the groups and missing data are related. This is usually performed in conjuction with an ANOVA which tests if the mean response (which varies with data type) is the same across groups; this combination is called IMD_ANOVA. It's probably a good idea to first filter the data with 'imd_anova_filter' to see if there is enough infomration to even do this test. See Webb-Robertson et al. (2010) for more.

Usage

imd_test(
  omicsData,
  groupData,
  comparisons,
  pval_adjust_multcomp,
  pval_adjust_fdr,
  pval_thresh,
  covariates,
  paired,
  parallel = TRUE
)

Arguments

Value

a list of 'data.frame's

Author(s)

Bryan Stanfill

References

Webb-Robertson, Bobbie-Jo M., et al. "Combined statistical analyses of peptide intensities and peptide occurrences improves identification of significant peptides from MS-based proteomics data." Journal of proteome research 9.11 (2010): 5748-5756.

Description

This function returns an imdanovaFilt object for use with applyFilt

Usage

imdanova_filter(omicsData)

Arguments

Details

The output from this function can be used in conjunction with applyFilt to filter out molecules that are not present in enough samples to do statistical comparisons. If any singleton groups are present in the omicsData object, those groups are not part of the filter object that is returned.

Value

An S3 object of class imdanovaFilt (also a data.frame) containing the molecule identifier and number of samples in each group with non-missing values for that molecule.

Author(s)

Kelly Stratton

Examples

library(pmartRdata)
mypep <- group_designation(omicsData = pep_object, main_effects = "Phenotype")
to_filter <- imdanova_filter(omicsData = mypep)
summary(to_filter, min_nonmiss_anova = 2)

Description

Select biomolecules for normalization via the method of the top L order statistics (LOS)

Usage

los(e_data, edata_id, L = 0.05)

Arguments

Details

The biomolecule abundances of the top L order statistics are identified and returned. Specifically, for each sample, the biomolecules with the top L proportion of highest absolute abundance are retained, and the union of these biomolecules is taken as the subset identified.

Value

Character vector containing the biomolecules belonging to the subset.

Author(s)

Kelly Stratton, Lisa Bramer

Description

Calculate normalization parameters for the data via median absolute deviation (MAD) transformation.

Usage

mad_transform(
  e_data,
  edata_id,
  subset_fn,
  feature_subset,
  backtransform = FALSE,
  apply_norm = FALSE,
  check.names = NULL
)

Arguments

Details

Each feature is scaled by subtracting the median of the feature subset specified for normalization and then dividing the result by the median absolute deviation (MAD) of the feature subset specified for normalization to get the normalized data. The location estimates are the subset medians for each sample. The scale estimates are the subset MADs for each sample. Medians are taken ignoring any NA values. If backtransform is TRUE, the normalized feature values are multiplied by a pooled MAD (estimated from all samples) and a global median of the subset data (across all samples) is added back to the normalized values.

Value

List containing two elements: norm_params is list with two elements:

backtransform_params is a list with two elements:

If backtransform is set to TRUE then each list item under backtransform_params will be NULL.

If apply_norm is TRUE, the transformed data is returned as a third list item.

Author(s)

Lisa Bramer, Kelly Stratton

Description

Calculate normalization parameters for the data via via mean centering.

Usage

mean_center(
  e_data,
  edata_id,
  subset_fn,
  feature_subset,
  backtransform = FALSE,
  apply_norm = FALSE,
  check.names = NULL
)

Arguments

Details

The sample-wise mean of the feature subset specified for normalization is subtracted from each feature in e_data to get the normalized data. The location estimates are the sample-wise means of the subset data. There are no scale estimates for mean centering, though the function returns a NULL list element as a placeholdfer for a scale estimate. If backtransform is TRUE, the global median of the subset data (across all samples) is added back to the normalized values. Medians are taken ignoring any NA values.

Value

List containing two elements: norm_params is list with two elements:

backtransform_params is a list with two elements:

If backtransform is set to TRUE then each list item under backtransform_params will be NULL.

If apply_norm is TRUE, the transformed data is returned as a third list item.

Author(s)

Lisa Bramer, Kelly Stratton

Description

Calculate normalization parameters for the data via median centering.

Usage

median_center(
  e_data,
  edata_id,
  subset_fn,
  feature_subset,
  backtransform = FALSE,
  apply_norm = FALSE,
  check.names = NULL
)

Arguments

Details

The sample-wise median of the feature subset specified for normalization is subtracted from each feature in e_data to get the normalized data. The location estimates are the sample-wise medians of the subset data. There are no scale estimates for median centering, though the function returns a NULL list element as a placeholder for a scale estimate. If backtransform is TRUE, the global median of the subset data (across all samples) is added back to the normalized values. Medians are taken ignoring any NA values.

Value

List containing two elements: norm_params is list with two elements:

backtransform_params is a list with two elements:

If backtransform is set to TRUE then each list item under backtransform_params will be NULL.

If apply_norm is TRUE, the transformed data is returned as a third list item.

Author(s)

Lisa Bramer, Kelly Stratton

Description

This function takes in an omicsData object, and outputs a list of two data frames, one containing the number of missing values by sample, and the other containing the number of missing values by molecule

Usage

missingval_result(omicsData)

Arguments

Value

S3 object of class naRes, which is a list of two data frames, one containing the number of missing values per sample, and the other containing the number of missing values per molecule. For count data, zeroes represent missing values; for abundance data, NA's represent missing values. This object can be used with 'plot' and 'summary' methods to examine the missing values in the dataset.

Examples

library(pmartRdata)
result1 = missingval_result(omicsData = lipid_neg_object)
result2 = missingval_result(omicsData = metab_object)

Description

This function returns a moleculeFilt object for use with applyFilt

Usage

molecule_filter(omicsData, use_groups = FALSE, use_batch = FALSE)

Arguments

Details

Attribute of molecule_filt object is "total_poss_obs", the number of total possible observations for each feature (same as the number of samples)

Value

An S3 object of class 'moleculeFilt' (also a data.frame) that contains the molecule identifier and the number of samples for which the molecule was observed (i.e. not NA)

Author(s)

Kelly Stratton

Examples

library(pmartRdata)
to_filter <- molecule_filter(omicsData = pep_object)
summary(to_filter, min_num = 2)

Description

This function computes the number of non-missing observations for samples, based on a group designation, for every biomolecule in the dataset

Usage

nonmissing_per_group(omicsData)

Arguments

Value

a list of length two. The first element giving the total number of possible samples for each group. The second element giving a data.frame with the first column giving the peptide and the second through kth columns giving the number of non-missing observations for each of the k groups.

Author(s)

Lisa Bramer, Kelly Stratton

Description

Calculates normalization parameters based on the data using the specified subset and normalization functions with option to apply the normalization to the data.

Usage

normalize_global(
  omicsData,
  subset_fn,
  norm_fn,
  params = NULL,
  apply_norm = FALSE,
  backtransform = FALSE,
  min_prop = NULL,
  check.names = NULL
)

Arguments

Details

Below are details for specifying function and parameter options.

Value

If apply_norm is FALSE, an S3 object of type 'normRes' is returned. This object contains a list with: subset method, normalization method, normalization parameters, number of biomolecules used in normalization, and proportion of biomolecules used in normalization. plot() and summary() methods are available for this object. If apply_norm is TRUE, then the normalized data is returned in an object of the appropriate S3 class (e.g. pepData).

Subset Functions

Specifying a subset function indicates the subset of biomolecules (rows of e_data) that should be used for computing normalization factors. The following are valid options: "all", "los", "ppp", "complete", "rip", and "ppp_rip". The option "all" is the subset that includes all biomolecules (i.e. no subsetting is done). The option "los" identifies the subset of the biomolecules associated with the top L order statistics, where L is a proportion between 0 and 1. Specifically, the biomolecules falling within the top L proportion of highest absolute abundance are retained for each sample, and the union of these biomolecules is taken as the subset identified (Wang et al., 2006). The option "ppp" (originally stands for percentage of peptides present) identifies the subset of biomolecules that are present/non-missing for a minimum proportion of samples (Karpievitch et al., 2009; Kultima et al., 2009). The option "complete" retains molecules with no missing data across all samples, equivalent to "ppp" with proportion = 1. The option "rip" identifies biomolecules with complete data that have a p-value greater than a defined threshold alpha (common values include 0.1 or 0.25) when subjected to a Kruskal-Wallis test based (non-parametric one-way ANOVA) on group membership (Webb-Robertson et al., 2011). The option "ppp_rip" is equivalent to "rip" however rather than requiring biomolecules with complete data, biomolecules with at least a proportion of non-missing values are subject to the Kruskal-Wallis test.

Normalization Functions

Specifying a normalization function indicates how normalization scale and location parameters should be calculated. The following are valid options: "median", "mean", "zscore", and "mad". For median centering, the location estimates are the sample-wise medians of the subset data and there are no scale estimates. For mean centering, the location estimates are the sample-wise means of the subset data and there are no scale estimates. For z-score transformation, the location estimates are the subset means for each sample and the scale estimates are the subset standard deviations for each sample. For median absolute deviation (MAD) transformation, the location estimates are the subset medians for each sample and the scale estimates are the subset MADs for each sample.

Specifying Subset Parameters Using the params Argument

Parameters for the chosen subset function should be specified in a list with the function specification followed by an equal sign and the desired parameter value. For example, if LOS with 0.1 is desired, one should use params = list(los = 0.1). ppp_rip can be specified in one of two ways: specify the parameters with each separate function or combine using a nested list (e.g. params = list(ppp_rip = list(ppp = 0.5, rip = 0.2))).

The following functions have parameters that can be specified:

Backtransform

The purpose of back transforming data is to ensure values are on a scale similar to their raw values before normaliztion. The following values are calculated and/or applied for backtransformation purposes:

Author(s)

Lisa Bramer

References

Webb-Robertson BJ, Matzke MM, Jacobs JM, Pounds JG, Waters KM. A statistical selection strategy for normalization procedures in LC-MS proteomics experiments through dataset-dependent ranking of normalization scaling factors. Proteomics. 2011;11(24):4736-41.

Examples

library(pmartRdata)

mymetab <- edata_transform(
  omicsData = metab_object,
  data_scale = "log2"
)
mymetab <- group_designation(
  omicsData = mymetab,
  main_effects = "Phenotype"
)
norm_object <- normalize_global(
  omicsData = mymetab,
  subset_fn = "all",
  norm_fn = "median"
)
norm_data <- normalize_global(
  omicsData = mymetab,
  subset_fn = "all",
  norm_fn = "median",
  apply_norm = TRUE,
  backtransform = TRUE
)

Description

This function is intended to be used in SPANS only. It is a VERY trimmed down version of normalize_global. It is trimmed down because within SPANS we only need the norm_params element from the output of the normalize_global function. All of the other options and output can be ignored.

Usage

normalize_global_basic(edata, norm_fn)

Arguments

Value

A list containing the location and scale parameters for normalizing the data.

Description

Examine reference pool samples and apply normalization of study samples to their corresponding reference pool sample

Usage

normalize_isobaric(
  omicsData,
  exp_cname = NULL,
  apply_norm = FALSE,
  channel_cname = NULL,
  refpool_channel = NULL,
  refpool_cname = NULL,
  refpool_notation = NULL
)

Arguments

Details

There are two ways to specify the information needed for identifying reference samples which should be used for normalization:

1. specify channel_cname and refpool_channel. This should be used when the reference sample for each experiment/plate was always located in the same channel. Here channel_cname gives the column name for the column in f_data which gives information about which channel each sample was run on, and refpool_channel is a character string specifying the value in channel_colname that corresponds to the reference sample channel.

2. specify refpool_cname and refpool_notation. This should be used when the reference sample is not in a consistent channel across experiments/plates. Here, refpool_cname gives the name of the column in f_data which indicates whether a sample is a reference or not, and refpool_notation is a character string giving the value used to denote a reference sample in that column.

In both cases you must specify exp_cname which gives the column name for the column in f_data containing information about which experiment/plate a sample was run on.

Value

If apply_norm = TRUE, an object of class 'isobaricpepData', normalized to reference pool, and with the attribute 'isobaric_info' updated to include information about the reference pool samples and the normalization procedure. Otherwise an object of class 'isobaricnormRes' containing similar information about the normalization process

Examples

library(pmartRdata)

myiso <- edata_transform(isobaric_object, "log2")

# Don't apply the normalization quite yet;
# can use summary() and plot() to view reference pool samples
myiso_refpools <- normalize_isobaric(
  omicsData = myiso, exp_cname = "Plex",
  apply_norm = FALSE,
  refpool_cname = "Virus",
  refpool_notation = "Pool"
)
summary(myiso_refpools)

# Now apply the normalization;
# can use plot() to view the study samples after reference pool normalization
myiso_norm <- normalize_isobaric(
  omicsData = myiso, exp_cname = "Plex",
  apply_norm = TRUE,
  refpool_cname = "Virus",
  refpool_notation = "Pool"
)

Description

Perform Loess normalization

Usage

normalize_loess(omicsData, method = "fast", span = 0.4)

Arguments

Details

A wrapper for the normalizeCyclicLoess function from the limma package.

Value

The normalized data is returned in an object of the appropriate S3 class (e.g. pepData), on the same scale as omicsData (e.g. if omicsData contains log2 transformed data, the normalization will be performed on the non-log2 scale and then re-scaled after normalization to be returned on the log2 scale).

References

Bolstad, B. M., Irizarry R. A., Astrand, M., and Speed, T. P. (2003). A comparison of normalization methods for high density oligonucleotide array data based on bias and variance. Bioinformatics 19, 185-193.

Ballman, KV Grill, DE, Oberg, AL and Therneau, TM (2004). Faster cyclic loess: normalizing RNA arrays via linear models. Bioinformatics 20, 2778-2786.

See Also

normalizeCyclicLoess in the limma package

Examples

library(pmartRdata)
mypep <- edata_transform(pep_object, "log2")
result <- normalize_loess(mypep)

Description

The data is normalized either to a spiked-in metabolite or to a sample-specific property

Usage

normalize_nmr(
  omicsData,
  apply_norm = FALSE,
  backtransform = FALSE,
  metabolite_name = NULL,
  sample_property_cname = NULL
)

Arguments

Details

There are two ways to specify the information needed for performing instrument normalization on an nmrData object:

1. specify metabolite_name. This should be used when normalization to a spiked in standard is desired. Here metabolite_name gives the name of the metabolite in e_data (and e_meta, if present) corresponding to the spiked in standard. If any samples have a missing value for this metabolite, an error is returned.

2. specify sample_property_cname. This should be used when normalizing to a sample property, such as concentration, is desired. Here, sample_property_cname gives the name of the column in f_data which contains the property to use for normalization. If any samples have a missing value for this column, and error is returned.

Value

If apply_norm is TRUE, an object of class 'nmrData', normalized and with information about the normalization process in 'nmr_info'. Otherwise, an object of class 'nmrnormRes' is returned, with the same info about normalization in attribute 'nmr_info' to be passed to plotting and summary functions.

Backtransform

The purpose of back transforming data is to ensure values are on a scale similar to their raw values before normalization. The following values are calculated and/or applied for backtransformation purposes:

See examples below.

Examples

library(pmartRdata)

# Normalize using a metabolite (this is merely an example of how to use this specification;
# the metabolite used was not actually spiked-in for the purpose of normalization)
mynmr <- edata_transform(
  omicsData = nmr_identified_object,
  data_scale = "log2"
)
nmr_norm <- normalize_nmr(
  omicsData = mynmr, apply_norm = TRUE,
  metabolite_name = "unkm1.53",
  backtransform = TRUE
)

# Normalization using a sample property
mynmr <- edata_transform(
  omicsData = nmr_identified_object,
  data_scale = "log2"
)
nmr_norm <- normalize_nmr(
  omicsData = mynmr, apply_norm = TRUE,
  sample_property_cname = "Concentration",
  backtransform = TRUE
)

Description

Perform quantile normalization

Usage

normalize_quantile(omicsData)

Arguments

Details

Quantile normalization is an algorithm for normalizing a set of data vectors by giving them the same distribution. It is applied to data on the abundance scale (e.g. not a log scale). It is often used for microarray data.

The method is implemented as described in Bolstad et al. (2003).

Value

The normalized data is returned in an object of the appropriate S3 class (e.g. pepData), on the same scale as omicsData (e.g. if omicsData contains log2 transformed data, the normalization will be performed on the non-log2 scale and then re-scaled after normalization to be returned on the log2 scale).

Author(s)

Kelly Stratton

References

Bolstad, B. M., Irizarry, R. A., Åstrand, M., & Speed, T. P. (2003). A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bioinformatics, 19(2), 185-193.