| Title: | Detecting regions of differential abundance between scRNA-seq datasets |
|---|---|
| Description: | DA-seq is a multiscale approach for detecting DA subpopulations. In contrast to clustering based approaches, our method can detect DA subpopulations that do not form well separated clusters. For each cell, we compute a multiscale differential abundance score measure. These scores are based on the k nearest neighbors in the transcriptome space for various values of k. |
| Authors: | Jun Zhao <[email protected]> |
| Maintainer: | Jun Zhao <[email protected]> |
| License: | |
| Version: | 1.0.0 |
| Built: | 2024-09-13 07:43:13 UTC |
| Source: | https://github.com/KlugerLab/DAseq |
Add DA region information to the meta.data of a Seurat object
addDAslot(object, da.regions, da.slot = "da", set.ident = F)addDAslot(object, da.regions, da.slot = "da", set.ident = F)
object |
input Seurat object |
da.regions |
output from function getDAregion() |
da.slot |
character, variable name to put in Seurat meta.data, default "da" |
set.ident |
a logical value to indicate whether to set Idents of the Seurat object to DA information, default False |
updated Seurat object
DA-seq is a multiscale approach for detecting DA subpopulations. In contrast to clustering based approaches, our method can detect DA subpopulations that do not form well separated clusters. For each cell, we compute a multiscale differential abundance score measure. These scores are based on the k nearest neighbors in the transcriptome space for various values of k.
Jun Zhao, Ariel Jaffe, Henry Li, Ofir Lindenbaum, Xiuyuan Cheng, Yuval Kluger
Step 1: compute a multiscale score measure for each cell of its k-nearest-neighborhood for multiple values of k. Step 2: train a logistic regression classifier based on the multiscale score measure and retain cells that may reside in DA regions.
getDAcells(X, cell.labels, labels.1, labels.2, k.vector = NULL, save.knn = F, alpha = 0, k.folds = 10, n.runs = 5, n.rand = 2, pred.thres = NULL, do.plot = T, plot.embedding = NULL, size = 0.5)getDAcells(X, cell.labels, labels.1, labels.2, k.vector = NULL, save.knn = F, alpha = 0, k.folds = 10, n.runs = 5, n.rand = 2, pred.thres = NULL, do.plot = T, plot.embedding = NULL, size = 0.5)
X |
size N-by-p matrix, input merged dataset of interest after dimension reduction. |
cell.labels |
size N character vector, labels for each input cell |
labels.1 |
character vector, label name(s) that represent condition 1 |
labels.2 |
character vector, label name(s) that represent condition 2 |
k.vector |
vector, k values to create the score vector |
save.knn |
a logical value to indicate whether to save computed kNN result, default False |
alpha |
numeric, elasticnet mixing parameter passed to glmnet(), default 0 (Ridge) |
k.folds |
integer, number of data splits used in the neural network, default 10 |
n.runs |
integer, number of times to run the neural network to get the predictions, default 5 |
n.rand |
integer, number of random permutations to run, default 2 |
pred.thres |
length-2 vector, top and bottom threshold on DA measure, default NULL, select significant DA cells based on permutation |
do.plot |
a logical value to indicate whether to return ggplot objects showing the results, default True |
plot.embedding |
size N-by-2 matrix, 2D embedding for the cells |
size |
cell size to use in the plot, default 0.5 |
a list of results
index of cells used in DA calculation
score vector for each cell
(mean) prediction from the logistic regression
index for DA cells more abundant in condition of labels.2
index for DA cells more abundant in condition of labels.1
ggplot object showing the predictions of logistic regression on plot.embedding
ggplot object highlighting cells of da.cell.idx on plot.embedding
Cluster the DA cells retained from Step 1 and Step 2 of DA-seq to obtain spatially coherent DA regions.
getDAregion(X, da.cells, cell.labels, labels.1, labels.2, prune.SNN = 1/15, resolution = 0.05, group.singletons = F, min.cell = NULL, do.plot = T, plot.embedding = NULL, size = 0.5, do.label = F, ...)getDAregion(X, da.cells, cell.labels, labels.1, labels.2, prune.SNN = 1/15, resolution = 0.05, group.singletons = F, min.cell = NULL, do.plot = T, plot.embedding = NULL, size = 0.5, do.label = F, ...)
X |
size N-by-p matrix, input merged dataset of interest after dimension reduction |
da.cells |
output from getDAcells() or updateDAcells() |
cell.labels |
size N vector, labels for each input cell |
labels.1 |
vector, label name(s) that represent condition 1 |
labels.2 |
vector, label name(s) that represent condition 2 |
prune.SNN |
parameter for Seurat function FindNeighbors(), default 1/15 |
resolution |
parameter for Seurat function FindClusters(), default 0.05 |
group.singletons |
parameter for Seurat function FindClusters(), default True |
min.cell |
integer, number of cells below which a DA region will be removed as outliers, default NULL, use minimum k value in k-vector |
do.plot |
a logical value to indicate whether to return ggplot objects showing the results, default True |
plot.embedding |
size N-by-2 matrix, 2D embedding for the cells |
size |
cell size to use in the plot, default 0.5 |
do.label |
a logical value to indicate whether to label each DA region with text, default False |
... |
other parameters to pass to Seurat FindClusters() |
a list of results
index of cells used in DA calculation
DA region label for each cell from the whole dataset, '0' represents non-DA cells.
a table showing DA score and p-value for each DA region
ggplot object showing DA regions on plot.embedding
Produce a ggplot object with cells on 2D embedding, colored by given labels of each cell.
plotCellLabel(X, label, cell.col = NULL, size = 0.5, alpha = 1, shape = 16, do.label = T, label.size = 4, label.repel = F, label.plot = NULL)plotCellLabel(X, label, cell.col = NULL, size = 0.5, alpha = 1, shape = 16, do.label = T, label.size = 4, label.repel = F, label.plot = NULL)
X |
matrix, 2D embedding of each cell for the plot |
label |
vector, label for each cell |
cell.col |
string vector, color bar to use for cell labels, default ggplot default |
size |
numeric, dot size for each cell, default 0.5 |
alpha |
numeric between 0 to 1, dot opacity for each cell, default 1 |
do.label |
a logical value indicating whether to add text to mark each cell label |
label.size |
numeric, size of text labels, default 4 |
label.repel |
a logical value indicating whether to repel the labels to avoid overlapping, default False |
label.plot |
cell labels to add text annotations, default NULL, add text for all labels |
a ggplot object
Produce a ggplot object with cells on 2D embedding, colored by a given score for each cell.
plotCellScore(X, score, cell.col = c("blue", "white", "red"), size = 0.5, alpha = 1, shape = 16)plotCellScore(X, score, cell.col = c("blue", "white", "red"), size = 0.5, alpha = 1, shape = 16)
X |
matrix, 2D embedding of each cell for the plot |
score |
numeric vector, a single value to color each cell, continuous |
cell.col |
string vector, color bar to use for "score", defaul c("blue","white","red") |
size |
numeric, dot size for each cell, default 0.5 |
alpha |
numeric between 0 to 1, dot opacity for each cell, default 1 |
a ggplot object
Plot da site
plotDAsite(X, site.list, size = 0.5, cols = NULL)plotDAsite(X, site.list, size = 0.5, cols = NULL)
X |
matrix, 2D embedding of each cell for the plot |
site.list |
list, a list of cell indices for each site to plot |
size |
numeric, dot size for each cell, default 0.5 |
cols |
string vector, color bar to use for each site, default ggplot default |
Run STG to select a set of genes that separate cells with label.1 from label.2 (other labels)
runSTG(X, X.labels, label.1, label.2 = NULL, lambda = 1.5, n.runs = 5, return.model = T, python.use = "/usr/bin/python", GPU = "")runSTG(X, X.labels, label.1, label.2 = NULL, lambda = 1.5, n.runs = 5, return.model = T, python.use = "/usr/bin/python", GPU = "")
X |
matrix, normalized expression matrix of all cells in the dataset, genes are in rows, rownames must be gene names |
X.labels |
numeric vector, specify labels for each cell, must be 0 or 1 |
label.1 |
cell label to define markers for |
label.2 |
second cell label to for comparison, if NULL, use all other labels |
lambda |
numeric, regularization parameter that weights the number of selected genes, a larger lambda leads to fewer genes, default 1.5 |
n.runs |
integer, number of runs to run the model, default 5 |
return.model |
a logical value to indicate whether to return the actual model of STG |
python.use |
character string, the Python to use, default "/usr/bin/python" |
GPU |
which GPU to use, default ”, using CPU |
a list of results:
a list of data.frame with markers for each DA region
a numeric vector showing mean accuracy for each DA region
a list of model for each DA region, each model contains:
the model of STG of the final run
cell names/indices used to train the model
features used to train the model
the selected features of the final run
the linear prediction value for each cell from the model
Use Seurat FindMarkers() function to identify genes that distinguish a DA region from its local neighborhood
SeuratLocalMarkers(object, da.slot = "da", da.region.to.run, cell.label.slot, cell.label.to.run, ...)SeuratLocalMarkers(object, da.slot = "da", da.region.to.run, cell.label.slot, cell.label.to.run, ...)
object |
input Seurat object |
da.slot |
character, variable name that represents DA regions in Seurat meta.data, default "da" |
da.region.to.run |
numeric, which (single) DA region to find local markers for |
cell.label.slot |
character, variable name that represents cell labeling information in Seurat meta.data to combine with DA information |
cell.label.to.run |
cell label(s) that represent the local neiborhood for the input DA region |
... |
parameters passed to Seurat FindMarkers() function |
a data.frame of markers and statistics
Use Seurat FindMarkers() function to identify characteristic genes for DA regions
SeuratMarkerFinder(object, da.slot = "da", da.regions.to.run = NULL, ...)SeuratMarkerFinder(object, da.slot = "da", da.regions.to.run = NULL, ...)
object |
input Seurat object |
da.slot |
character, variable name that represents DA regions in Seurat meta.data, default "da" |
da.regions.to.run |
numeric (vector), which DA regions to find markers for, default is to run all regions |
... |
parameters passed to Seurat FindMarkers() function |
a list of markers for DA regions with statistics
STG local markers Run STG to find a set of genes that separate a given DA region from a local subset of cells.
STGlocalMarkers(X, da.regions, da.region.to.run, cell.label.info, cell.label.to.run, ...)STGlocalMarkers(X, da.regions, da.region.to.run, cell.label.info, cell.label.to.run, ...)
X |
matrix, normalized expression matrix of all cells in the dataset, genes are in rows, rownames must be gene names |
da.regions |
output from the function getDAregion() |
da.region.to.run |
numeric, which (single) DA region to find local markers for |
cell.label.info |
vector, cell labeling information to select the local subset of cells to compare with input DA region |
cell.label.to.run |
cell label(s) to select from cell.label.info that represent the local neiborhood for the input DA region |
lambda |
numeric, regularization parameter that weights the number of selected genes, a larger lambda leads to fewer genes, default 1.5 |
n.runs |
integer, number of runs to run the model, default 5 |
python.use |
character string, the Python to use, default "/usr/bin/python" |
return.model |
a logical value to indicate whether to return the actual model of STG |
GPU |
which GPU to use, default ”, using CPU |
a list of results:
a list of data.frame with markers for each DA region
a numeric vector showing mean accuracy for each DA region
a list of model for each DA region, each model contains:
the model of STG of the final run
features used to train the model
the selected features of the final run
the linear prediction value for each cell from the model
Use STG (stochastic gates) to select genes that separate each DA region from the rest of the cells. For a full description of the algorithm, see Y. Yamada, O. Lindenbaum, S. Negahban, and Y. Kluger. Feature selection using stochastic gates. arXiv preprint arXiv:1810.04247, 2018.
STGmarkerFinder(X, da.regions, da.regions.to.run = NULL, lambda = 1.5, n.runs = 5, return.model = T, python.use = "/usr/bin/python", GPU = "")STGmarkerFinder(X, da.regions, da.regions.to.run = NULL, lambda = 1.5, n.runs = 5, return.model = T, python.use = "/usr/bin/python", GPU = "")
X |
matrix, normalized expression matrix of all cells in the dataset, genes are in rows, rownames must be gene names |
da.regions |
output from the function getDAregion() |
da.regions.to.run |
numeric (vector), which DA regions to run the marker finder, default is to run all regions |
lambda |
numeric, regularization parameter that weights the number of selected genes, a larger lambda leads to fewer genes, default 1.5 |
n.runs |
integer, number of runs to run the model, default 5 |
return.model |
a logical value to indicate whether to return the actual model of STG |
python.use |
character string, the Python to use, default "/usr/bin/python" |
GPU |
which GPU to use, default ”, using CPU |
a list of results:
a list of data.frame with markers for each DA region
a numeric vector showing mean accuracy for each DA region
a list of model for each DA region, each model contains:
the model of STG of the final run
features used to train the model
the selected features of the final run
the linear prediction value for each cell from the model
Use different threshold to select DA cells based on an output from getDAcells().
updateDAcells(X, pred.thres = NULL, force.thres = F, alpha = NULL, k.folds = 10, n.runs = 10, cell.labels = NULL, labels.1 = NULL, labels.2 = NULL, do.plot = T, plot.embedding = NULL, size = 0.5)updateDAcells(X, pred.thres = NULL, force.thres = F, alpha = NULL, k.folds = 10, n.runs = 10, cell.labels = NULL, labels.1 = NULL, labels.2 = NULL, do.plot = T, plot.embedding = NULL, size = 0.5)
X |
output from getDAcells() |
pred.thres |
length-2 vector, top and bottom threshold on DA measure, default NULL, select significant DA cells based on permutation |
force.thres |
a logical value to indicate whether to forcefully use pred.thres without considering significance, default False |
alpha |
set this parameter to not NULL to rerun Logistic regression |
do.plot |
a logical value to indicate whether to return ggplot objects showing the results, default True |
plot.embedding |
size N-by-2 matrix, 2D embedding for the cells |
size |
cell size to use in the plot, default 0.5 |
a list of results with updated DA cells
A matrix containing 2D t-SNE embedding of the melanoma dataset
X.2d.melanomaX.2d.melanoma
An object of class matrix with 16291 rows and 2 columns.
https://www.sciencedirect.com/science/article/pii/S0092867418313941 Sade-Feldman, Moshe, et al. (Cell. 2018)
A dataset containing information of each sample, indicating whether this sample is a "responder" (R) or a "non-responder" (NR)
X.label.infoX.label.info
a data.frame with 48 rows and 2 columns
sample label, matching with labels in X.label.melanoma
condition of the sample label, either R or NR
https://www.sciencedirect.com/science/article/pii/S0092867418313941 Sade-Feldman, Moshe, et al. (Cell. 2018)
A string vector with the length equal to number of cells, indicating sample labels of each cell: which sample each cell comes from
X.label.melanomaX.label.melanoma
An object of class character of length 16291.
https://www.sciencedirect.com/science/article/pii/S0092867418313941 Sade-Feldman, Moshe, et al. (Cell. 2018)
A dataset containing the top 10 PCs (principal components) of the melanoma dataset
X.melanomaX.melanoma
An object of class matrix with 16291 rows and 10 columns.
https://www.sciencedirect.com/science/article/pii/S0092867418313941 Sade-Feldman, Moshe, et al. (Cell. 2018)