| Title: | Imputation of label-free mass spectrometry peptides |
|---|---|
| Description: | MsImpute is a package for imputation of peptide intensity in proteomics experiments. It additionally contains tools for MAR/MNAR diagnosis and assessment of distortions to the probability distribution of the data post imputation. The missing values are imputed by low-rank approximation of the underlying data matrix if they are MAR (method = "v2"), by Barycenter approach if missingness is MNAR ("v2-mnar"), or by Peptide Identity Propagation (PIP). |
| Authors: | Soroor Hediyeh-zadeh [aut, cre]
|
| Maintainer: | Soroor Hediyeh-zadeh <[email protected]> |
| License: | GPL (>=2) |
| Version: | 1.15.0 |
| Built: | 2024-07-20 04:57:17 UTC |
| Source: | https://github.com/bioc/msImpute |
For an imputed dataset, it computes within phenotype/experimental condition similarity (i.e. preservation of local structures), between phenotype distances (preservation of global structures), and the Gromov-Wasserstein (GW) distance between original (source) and imputed data.
computeStructuralMetrics(x, group = NULL, y = NULL, k = 2)computeStructuralMetrics(x, group = NULL, y = NULL, k = 2)
x |
numeric matrix. An imputed data matrix of log-intensity. |
group |
factor. A vector of biological groups, experimental conditions or phenotypes (e.g. control, treatment). |
y |
numeric matrix. The source data (i.e. the original log-intensity matrix),
preferably subsetted on highly variable peptides (see |
k |
numeric. Number of Principal Components used to compute the GW distance. default to 2. |
For each group of experimental conditions (e.g. treatment and control), the group centroid is
calculated as the average of observed peptide intensities. Withinness for each group is computed as
sum of the squared distances between samples in that group and
the group centroid. Betweenness is computed as sum of the squared distances between group centroids.
When comparing imputation approaches, the optimal imputation strategy should minimize the within
group distances, hence smaller withinness, and maximizes between group distances, hence larger betweenness.
The GW metric considers preservation of both local and global structures simultaneously. A small GW distance
suggests that imputation has introduced small distortions to global and local structures overall, whereas a
large distance implies significant distortions. When comparing two or more imputation methods, the optimal
method is the method with smallest GW distance. The GW distance is computed on Principal Components (PCs)
of the source and imputed data, instead of peptides. Principal components capture the geometry of the data,
hence GW computed on PCs is a better measure of preservation of local and global structures. The PCs in the
source data are recommended to be computed on peptides with high biological variance. Hence, users are
recommended to subset the source data only on highly variable peptides (hvp) (see findVariableFeatures).
Since the hvp peptides have high biological variance, they are likely to have enough information to discriminate
samples from different experimental groups. Hence, PCs computed on those peptides should be representative
of the original source data with missing values. If the samples cluster by experimental group in the first
couple of PCs, then a choice of k=2 is reasonable. If the desired separation/clustering of samples
occurs in later PCs (i.e. the first few PCs are dominated by batches or unwanted variability), then
it is recommended to use a larger number of PCs to compute the GW metric.
If you are interested in how well the imputed data represent the original data in all possible dimensions,
then set k to the number of samples in the data (i.e. the number of columns in the intensity matrix).
GW distance estimation requires python. See example. All metrics are on log scale.
list of three metrics: withinness (sum of squared distances within a phenotype group),
betweenness (sum of squared distances between the phenotypes), and gromov-wasserstein distance (if xna is not NULL).
if group is NULL only the GW distance is returned. All metrics are on log scale.
Hediyeh-zadeh, S., Webb, A. I., & Davis, M. J. (2020). MSImpute: Imputation of label-free mass spectrometry peptides by low-rank approximation. bioRxiv.
data(pxd010943) y <- log2(data.matrix(pxd010943)) y <- y[complete.cases(y),] group <- as.factor(gsub("_[1234]", "", colnames(y))) computeStructuralMetrics(y, group, y=NULL)data(pxd010943) y <- log2(data.matrix(pxd010943)) y <- y[complete.cases(y),] group <- as.factor(gsub("_[1234]", "", colnames(y))) computeStructuralMetrics(y, group, y=NULL)
Spearman correlation between pairwise distances in the original data and imputed data. CPD quantifies preservation of the global structure after imputation. Requires complete datasets - for developers/use in benchmark studies only.
CPD(xorigin, ximputed)CPD(xorigin, ximputed)
xorigin |
numeric matrix. The original log-intensity data. Can not contain missing values. |
ximputed |
numeric matrix. The imputed log-intensity data. Can not contain missing values. |
numeric
data(pxd007959) y <- pxd007959$y y <- y[complete.cases(y),] # for demonstration we use same y for xorigin and ximputed CPD(y, y)data(pxd007959) y <- pxd007959$y y <- y[complete.cases(y),] # for demonstration we use same y for xorigin and ximputed CPD(y, y)
Every Modified sequence - Charge is considered as a precursor feature.
Only the feature with maximum intensity is retained. The columns are run names, the rows
are peptide ids (in the Modified.sequence_Charge format)
evidenceToMatrix( evidence, run_id = "Raw.file", peptide_id = "PeptideID", return_EList = FALSE, weights = NULL )evidenceToMatrix( evidence, run_id = "Raw.file", peptide_id = "PeptideID", return_EList = FALSE, weights = NULL )
evidence |
data.frame. The evidence table read from evidence.txt, or data.frame created by |
run_id |
character. The name of the column of evidence containing the run/raw file name. These form the columns of the intensity data matrix. |
peptide_id |
character. The name of the column of evidence containing the peptide ids. These form the rows of the intensity data matrix. |
return_EList |
logical. If TRUE, returns a |
weights |
character. The name of the column of evidence containing weights from |
The EListRaw object created by the function is intended to bridge msImpute and statistical
methods of limma. The object can be passed to normalizeBetweenArrays for normalisation, which can then
be passed to lmFit and eBayes for fitting linear models per peptide and Empirical Bayes moderation of t-statistics
respectively. The weights slot is recognized by lmFit, which incorporates the uncertainty in intensity values
inferred by PIP into the test statistic.
The function is also a generic tool to create a matrix or limma-compatible objects from the evidence table of MaxQuant.
a numeric matrix of intensity data, or a EListRaw object containing
such data and observation-level weights from mspip.
Soroor Hediyeh-zadeh
mspip
For each peptide, the total variance is decomposed into biological and technical variance using package scran
findVariableFeatures(y)findVariableFeatures(y)
y |
numeric matrix giving log-intensity. Can contain NA values. Peptides with insufficient observations will be ignored. |
A loess trend is fitted to total sample variances and mean intensities. For each peptide, the biological variance is then computed by subtracting the estimated technical variance from the loess fit from the total sample variance.
A data frame where rows are peptides and columns contain estimates of biological and technical variances. Peptides are ordered by biological variance.
computeStructuralMetrics
data(pxd007959) findVariableFeatures(data.matrix(log2(pxd007959$y)))data(pxd007959) findVariableFeatures(data.matrix(log2(pxd007959$y)))
The fraction of k-nearest class means in the original data that are preserved as k-nearest class means in imputed data. KNC quantifies preservation of the mesoscopic structure after imputation. Requires complete datasets - for developers/use in benchmark studies only.
KNC(xorigin, ximputed, class, k = 3)KNC(xorigin, ximputed, class, k = 3)
xorigin |
numeric matrix. The original log-intensity data. Can contain missing values. |
ximputed |
numeric matrix. The imputed log-intensity data. |
class |
factor. A vector of length number of columns (samples) in the data specifying the class/label (i.e. experimental group) of each sample. |
k |
number of nearest class means. default to k=3. |
numeric The proportion of preserved k-nearest class means in imputed data.
data(pxd007959) y <- pxd007959$y y <- y[complete.cases(y),] # for demonstration we use same y for xorigin and ximputed KNC(y, y, class = as.factor(pxd007959$samples$group))data(pxd007959) y <- pxd007959$y y <- y[complete.cases(y),] # for demonstration we use same y for xorigin and ximputed KNC(y, y, class = as.factor(pxd007959$samples$group))
The fraction of k-nearest neighbours in the original data that are preserved as k-nearest neighbours in imputed data. KNN quantifies preservation of the local, or microscopic structure. Requires complete datasets - for developers/use in benchmark studies only.
KNN(xorigin, ximputed, k = 3)KNN(xorigin, ximputed, k = 3)
xorigin |
numeric matrix. The original log-intensity data. Can not contain missing values. |
ximputed |
numeric matrix. The imputed log-intensity data. Can not contain missing values. |
k |
number of nearest neighbours. default to k=3. |
numeric The proportion of preserved k-nearest neighbours in imputed data.
data(pxd007959) y <- pxd007959$y y <- y[complete.cases(y),] # for demonstration we use same y for xorigin and ximputed KNN(y, y)data(pxd007959) y <- pxd007959$y y <- y[complete.cases(y),] # for demonstration we use same y for xorigin and ximputed KNN(y, y)
Returns a completed matrix of peptide log-intensity where missing values (NAs) are imputated
by low-rank approximation of the input matrix. Non-NA entries remain unmodified. msImpute requires at least 4
non-missing measurements per peptide across all samples. It is assumed that peptide intensities (DDA), or MS1/MS2 normalised peak areas (DIA),
are log2-transformed and normalised (e.g. by quantile normalisation).
msImpute( y, method = c("v2-mnar", "v2", "v1"), group = NULL, a = 0.2, rank.max = NULL, lambda = NULL, thresh = 1e-05, maxit = 100, trace.it = FALSE, warm.start = NULL, final.svd = TRUE, biScale_maxit = 20, gauss_width = 0.3, gauss_shift = 1.8 )msImpute( y, method = c("v2-mnar", "v2", "v1"), group = NULL, a = 0.2, rank.max = NULL, lambda = NULL, thresh = 1e-05, maxit = 100, trace.it = FALSE, warm.start = NULL, final.svd = TRUE, biScale_maxit = 20, gauss_width = 0.3, gauss_shift = 1.8 )
y |
Numeric matrix giving log-intensity where missing values are denoted by NA. Rows are peptides, columns are samples. |
method |
character. Allowed values are |
group |
character or factor vector of length |
a |
numeric. the weight parameter. default to 0.2. Weights the MAR-imputed distribution in the imputation scheme. |
rank.max |
Numeric. This restricts the rank of the solution. is set to min(dim( |
lambda |
Numeric. Nuclear-norm regularization parameter. Controls the low-rank property of the solution to the matrix completion problem. By default, it is determined at the scaling step. If set to zero the algorithm reverts to "hardImputation", where the convergence will be slower. Applicable to "v1" only. |
thresh |
Numeric. Convergence threshold. Set to 1e-05, by default. Applicable to "v1" only. |
maxit |
Numeric. Maximum number of iterations of the algorithm before the algorithm is converged. 100 by default. Applicable to "v1" only. |
trace.it |
Logical. Prints traces of progress of the algorithm. Applicable to "v1" only. |
warm.start |
List. A SVD object can be used to initialize the algorithm instead of random initialization. Applicable to "v1" only. |
final.svd |
Logical. Shall final SVD object be saved? The solutions to the matrix completion problems are computed from U, D and V components of final SVD. Applicable to "v1" only. |
biScale_maxit |
number of iteration for the scaling algorithm to converge . See |
gauss_width |
numeric. The width parameter of the Gaussian distribution to impute the MNAR peptides (features). This the width parameter in the down-shift imputation method. |
gauss_shift |
numeric. The shift parameter of the Gaussian distribution to impute the MNAR peptides (features). This the width parameter in the down-shift imputation method. |
msImpute operates on the softImpute-als algorithm in softImpute package.
The algorithm estimates a low-rank matrix ( a smaller matrix
than the input matrix) that approximates the data with a reasonable accuracy. SoftImpute-als determines the optimal
rank of the matrix through the lambda parameter, which it learns from the data.
This algorithm is implemented in method="v1".
In v2 we have used a information theoretic approach to estimate the optimal rank, instead of relying on softImpute-als
defaults. Similarly, we have implemented a new approach to estimate lambda from the data. Low-rank approximation
is a linear reconstruction of the data, and is only appropriate for imputation of MAR data. In order to make the
algorithm applicable to MNAR data, we have implemented method="v2-mnar" which imputes the missing observations
as weighted sum of values imputed by msImpute v2 (method="v2") and random draws from a Gaussian distribution.
Missing values that tend to be missing completely in one or more experimental groups will be weighted more (shrunken) towards
imputation by sampling from a Gaussian parameterised by smallest observed values in the sample (similar to minProb, or
Perseus). However, if the missing value distribution is even across the samples for a peptide, the imputed values
for that peptide are shrunken towards
low-rank imputed values. The judgment of distribution of missing values is based on the EBM metric implemented in
selectFeatures, which is also a information theory measure.
Missing values are imputed by low-rank approximation of the input matrix. If input is a numeric matrix, a numeric matrix of identical dimensions is returned.
Soroor Hediyeh-zadeh
Hastie, T., Mazumder, R., Lee, J. D., & Zadeh, R. (2015). Matrix completion and low-rank SVD via fast alternating least squares. The Journal of Machine Learning Research, 16(1), 3367-3402.
Hediyeh-zadeh, S., Webb, A. I., & Davis, M. J. (2020). MSImpute: Imputation of label-free mass spectrometry peptides by low-rank approximation. bioRxiv.
selectFeatures
data(pxd010943) y <- log2(data.matrix(pxd010943)) group <- gsub("_[1234]","", colnames(y)) yimp <- msImpute(y, method="v2-mnar", group=group)data(pxd010943) y <- log2(data.matrix(pxd010943)) group <- gsub("_[1234]","", colnames(y)) yimp <- msImpute(y, method="v2-mnar", group=group)
Peptide identity (sequence and charge) is propagated from MS-MS or PASEF identified features in evidence.txt to
MS1 features in allPeptides.txt that are detected but not identified. A confidence score (probability)
is assigned to every propagation. The confidence scores can be used as observation-level weights
in limma::lmFit to account for uncertainty in inferred peptide intensity values.
mspip( path_txt, k = 10, thresh = 0, skip_weights = TRUE, tims_ms = FALSE, group_restriction = NULL, nlandmarks = 50 )mspip( path_txt, k = 10, thresh = 0, skip_weights = TRUE, tims_ms = FALSE, group_restriction = NULL, nlandmarks = 50 )
path_txt |
character. The path to MaxQuant |
k |
numeric. The |
thresh |
numeric. The uncertainty threshold for calling a Identity Transfer as confident. Sequence to peptide
feature assignments with confidence score (probability) above a threshold (specified by |
skip_weights |
logical. If TRUE, the propagation confidence scores are also reported.
The confidence scores can be used as observation-level weights in |
tims_ms |
logical. Is data acquired by TIMS-MS? default to FALSE. |
group_restriction |
A data.frame with two columns named Raw.file and group, specifying run file and the (experimental) group to which the run belongs. Use this option for Unbalanced PIP |
nlandmarks |
numeric. Number of landmark peptides used for measuring neighborhood/coelution similarity. Default to 50. |
Data completeness is maximised by Peptide Identity Propagation (PIP) from runs where
a peptide is identified by MSMS or PASEF to runs where peptide is not fragmented
(hence MS2 information is not available), but is detected at the MS1 level. mspip reports a
confidence score for each peptide that was identified by PIP. The intensity values of PIP peptides
can be used to reduce missing values, while the reported confidence scores can be used to
weight the contribution of these peptide intensity values to variance estimation in linear models fitted in
limma.
Soroor Hediyeh-zadeh
evidenceToMatrix
For each peptide, the squares of coefficient of variations are computed and plotted against average log-intensity. Additionally, a loess trend is fitted to the plotted values. Outlier observations (possibly originated from incorrect match between runs), are detected and highlighted. Users can use this plot as a diagnostic plot to determine if filtering by average intensity is required.
plotCV2(y, trend = TRUE, main = NULL, ...)plotCV2(y, trend = TRUE, main = NULL, ...)
y |
numeric matrix of log-intensity |
trend |
logical. Should a loess trend be fitted to CV^2 and mean values. Default to TRUE. |
main |
character string. Title of the plot. Default to NULL. |
... |
any parameter passed to |
Outliers are determined by computing the RBF kernels, which reflect the chance that an observed point belong to the dataset (i.e. is close enough in distance to other data points). Users can determine the cut-off for intensity-based filtering with respect to the mean log-intensity of the outlier points.
A plot is created on the current graphics device.
data(pxd010943) y <- pxd010943 y <- log2(y) ppCV2 <- plotCV2(y)data(pxd010943) y <- pxd010943 y <- log2(y) ppCV2 <- plotCV2(y)
Extracellular vesicles isolated from the descending colon of pediatric patients with inflammatory bowel disease and control patients. Characterizes the proteomic profile of extracellular vesicles isolated from the descending colon of pediatric patients with inflammatory bowel disease and control participants. This object contains data from peptide.txt table output by MaxQuant. Rows are Modified Peptide IDs. Charge state variations are treated as distinct peptide species. Reverse complements and contaminant peptides are discarded. Peptides with more than 4 observed intensity values are retained. Additionally, qualified peptides are required to map uniquely to proteins. Two of the samples with missing group annotation were excluded. The peptide.txt and experimentalDesignTemplate files can be downloaded as RDS object from https://github.com/soroorh/proteomicscasestudies. Code for data processing is provided in package vignette.
pxd007959pxd007959
A list of two: samples (data frame of sample descriptions), and y (numeric matrix of peptide intensity values)
http://proteomecentral.proteomexchange.org/cgi/GetDataset?ID=PXD007959
Zhang X, Deeke SA, Ning Z, Starr AE, Butcher J, Li J, Mayne J, Cheng K, Liao B, Li L, Singleton R, Mack D, Stintzi A, Figeys D, Metaproteomics reveals associations between microbiome and intestinal extracellular vesicle proteins in pediatric inflammatory bowel disease. Nat Commun, 9(1):2873(2018)
Contains Peak Area for peptides in PXD010943. This study investigates the proteomic alterations in bone marrow neutrophils isolated from 5-8 week old Gfi1+/-, Gfi1K403R/-, Gfi1R412X/-, and Gfi1R412X/R412X mice using the SWATH-MS technique. This dataset consists of 13 SWATH-DIA runs on a TripleTOF 5600 plus (SCIEX). Rows are peptides. Charge state variations are treated as distinct peptide species. Peptides with more than 4 observed intensity values are retained. The peptide.txt and experimentalDesignTemplate files can be downloaded as RDS object from https://github.com/soroorh/proteomicscasestudies. Code for data processing is provided in package vignette.
pxd010943pxd010943
A matrix
http://proteomecentral.proteomexchange.org/cgi/GetDataset?ID=PXD010943
Muench DE, Olsson A, Ferchen K, Pham G, Serafin RA, Chutipongtanate S, Dwivedi P, Song B, Hay S, Chetal K, Trump-Durbin LR, Mookerjee-Basu J, Zhang K, Yu JC, Lutzko C, Myers KC, Nazor KL, Greis KD, Kappes DJ, Way SS, Salomonis N, Grimes HL, Mouse models of neutropenia reveal progenitor-stage-specific defects. Nature, 582(7810):109-114(2020)
A Trapped Ion Mobility Spectrometry (TIMS) dataset of blood plasma from a number of patients acquired in two batches. This is a technical dataset published by MaxQuant to benchmark their software for ion mobility enhanced shotgun proteomics. Rows are Modified Peptide IDs. Charge state variations are treated as distinct peptide species. For peptides with multiple identification types, the intensity is considered to be the median of reported intensity values. Reverse complememts and contaminant peptides are discarded. Peptides with more than 4 observed intensity values are retained. This object contains data from peptide.txt table output by MaxQuant. The evidence.txt file can be downloaded as RDS object from https://github.com/soroorh/proteomicscasestudies. Code for data processing is provided in package vignette.
pxd014777pxd014777
A matrix
http://proteomecentral.proteomexchange.org/cgi/GetDataset?ID=PXD014777
Prianichnikov N, Koch H, Koch S, Lubeck M, Heilig R, Brehmer S, Fischer R, Cox J, MaxQuant Software for Ion Mobility Enhanced Shotgun Proteomics. Mol Cell Proteomics, 19(6):1058-1069(2020)
Standardize a matrix to have optionally row means zero and variances one, and/or column means zero and variances one.
scaleData( object, maxit = 20, thresh = 1e-09, row.center = TRUE, row.scale = TRUE, col.center = TRUE, col.scale = TRUE, trace = FALSE )scaleData( object, maxit = 20, thresh = 1e-09, row.center = TRUE, row.scale = TRUE, col.center = TRUE, col.scale = TRUE, trace = FALSE )
object |
numeric matrix giving log-intensity where missing values are denoted by NA. Rows are peptides, columns are samples. |
maxit |
numeric. maximum iteration for the algorithm to converge (default to 20). When both row and column centering/scaling is requested, iteration may be necessary. |
thresh |
numeric. Convergence threshold (default to 1e-09). |
row.center |
logical. if row.center==TRUE (the default), row centering will be performed resulting in a matrix with row means zero. If row.center is a vector, it will be used to center the rows. If row.center=FALSE nothing is done. |
row.scale |
if row.scale==TRUE, the rows are scaled (after possibly centering, to have variance one. Alternatively, if a positive vector is supplied, it is used for row centering. |
col.center |
Similar to row.center |
col.scale |
Similar to row.scale |
trace |
logical. With trace=TRUE, convergence progress is reported, when iteration is needed. |
Standardizes rows and/or columns of a matrix with missing values, according to the biScale algorithm in Hastie et al. 2015.
Data is assumed to be normalised and log-transformed. Please note that data scaling might not be appropriate for MS1 data. A good strategy
is to compare mean-variance plot (plotCV2) before and after imputation. If the plots look differently, you may need to skip
data scaling. The MS1 data are more variable (tend to have higher CV^2), and may contain outliers which will skew the scaling.
A list of two components: E and E.scaled. E contains the input matrix, E.scaled contains the scaled data
Hastie, T., Mazumder, R., Lee, J. D., & Zadeh, R. (2015). Matrix completion and low-rank SVD via fast alternating least squares. The Journal of Machine Learning Research, 16(1), 3367-3402.
Hediyeh-zadeh, S., Webb, A. I., & Davis, M. J. (2020). MSImpute: Imputation of label-free mass spectrometry peptides by low-rank approximation. bioRxiv.
selectFeatures, msImpute
data(pxd010943) y <- pxd010943 y <- log2(y) keep <- (rowSums(!is.na(y)) >= 4) y <- as.matrix.data.frame(y[keep,]) y <- scaleData(y, maxit=30)data(pxd010943) y <- pxd010943 y <- log2(y) keep <- (rowSums(!is.na(y)) >= 4) y <- as.matrix.data.frame(y[keep,]) y <- scaleData(y, maxit=30)
Two methods are provided to identify features (peptides or proteins) that can be informative of missing patterns.
Method hvp fits a linear model to peptide dropout rate (proportion of samples were peptide is missing)
against peptide abundance (average log2-intensity). Method emb is a information theoretic approach to
identify missing patterns. It quantifies the heterogeneity (entropy) of missing patterns per
biological (experimental group). This is the default method.
selectFeatures( x, method = c("ebm", "hvp"), group, n_features = 500, suppress_plot = TRUE )selectFeatures( x, method = c("ebm", "hvp"), group, n_features = 500, suppress_plot = TRUE )
x |
Numeric matrix giving log-intensity where missing values are denoted by NA. Rows are peptides, columns are samples. |
method |
character. What method should be used to find features? options include |
group |
character or factor vector specifying biological (experimental) group e.g. control, treatment, WT, KO |
n_features |
Numeric, number of features with high dropout rate. 500 by default. Applicable if |
suppress_plot |
Logical show plot of dropouts vs abundances. Default to TRUE. Applicable if |
In general, the presence of group-wise (structured) blocks of missing values,
where peptides are missing in one experimental group can indicate MNAR, whereas if
such patterns are absent (or missingness is uniform across the samples), peptides are likely MAR.
In the presence of MNAR, left-censored MNAR imputation methods should
be chosen. Two methods are provided to explore missing patterns: method=hvp identifies top n_features
peptides with high average expression that also have high dropout rate, defined as the proportion of samples where
peptide is missing. Peptides with high (potentially) biological dropouts are marked in the hvp column in the
output dataframe. This method does not use any information about experimental conditions (i.e. group).
Another approach to explore and quantify missing patterns is by looking at how homogeneous or heterogeneous
missing patterns are in each experimental group. This is done by computing entropy of distribution of observed values.
This is the default and recommended method for selectFeatures. Entropy is reported in EBM column
of the output. A NaN EBM indicates peptide is missing at least in one experimental group. Features set to
TRUE in msImpute_feature column are the features selected by the selected method. Users are encouraged
to use the EBM metric to find informative features, hence why the group argument is required.
A data frame with a logical column denoting the selected features
Soroor Hediyeh-zadeh
Hediyeh-zadeh, S., Webb, A. I., & Davis, M. J. (2020). MSImpute: Imputation of label-free mass spectrometry peptides by low-rank approximation. bioRxiv.
msImpute
data(pxd007959) group <- pxd007959$samples$group y <- data.matrix(pxd007959$y) y <- log2(y) hdp <- selectFeatures(y, method="ebm", group = group) # construct matrix M to capture missing entries M <- ifelse(is.na(y),1,0) M <- M[hdp$msImpute_feature,] # plot a heatmap of missingness patterns for the selected peptides require(ComplexHeatmap) hm <- Heatmap(M, column_title = "dropout pattern, columns ordered by dropout similarity", name = "Intensity", col = c("#8FBC8F", "#FFEFDB"), show_row_names = FALSE, show_column_names = TRUE, cluster_rows = TRUE, cluster_columns = TRUE, show_column_dend = TRUE, show_row_dend = FALSE, row_names_gp = gpar(fontsize = 7), column_names_gp = gpar(fontsize = 8), heatmap_legend_param = list(#direction = "horizontal", heatmap_legend_side = "bottom", labels = c("observed","missing"), legend_width = unit(6, "cm")), ) hm <- draw(hm, heatmap_legend_side = "left")data(pxd007959) group <- pxd007959$samples$group y <- data.matrix(pxd007959$y) y <- log2(y) hdp <- selectFeatures(y, method="ebm", group = group) # construct matrix M to capture missing entries M <- ifelse(is.na(y),1,0) M <- M[hdp$msImpute_feature,] # plot a heatmap of missingness patterns for the selected peptides require(ComplexHeatmap) hm <- Heatmap(M, column_title = "dropout pattern, columns ordered by dropout similarity", name = "Intensity", col = c("#8FBC8F", "#FFEFDB"), show_row_names = FALSE, show_column_names = TRUE, cluster_rows = TRUE, cluster_columns = TRUE, show_column_dend = TRUE, show_row_dend = FALSE, row_names_gp = gpar(fontsize = 7), column_names_gp = gpar(fontsize = 8), heatmap_legend_param = list(#direction = "horizontal", heatmap_legend_side = "bottom", labels = c("observed","missing"), legend_width = unit(6, "cm")), ) hm <- draw(hm, heatmap_legend_side = "left")