mkarikom / rsoptsc

Cell-cell communication and lineage inference for scRNA-seq data

RSoptSC

Updates

• Demo SoptSC at vignette("Joost_et_al")

• Read the SoptSC article in NAR

• SoptSC is also available as a MATLAB package here

Installation

install.packages("devtools")
library(devtools)
install_github("mkarikom/RSoptSC")

Features

• Inference of cell-cell communication between single cells

• Integration of multiple analyses in a consistent mathematical framework: clustering, marker genes, pseudotime, and lineage inference

• Cell-cell similarity matrix construction to improve clustering

• NMF-based marker gene identification

• Prediction of the number of clusters present (via eigengap properties of the similarity matrix)

• Prediction of the initial cell in pseudotime

Documentation

Full details of RSoptSC and examples are available here.

rsoptsc's Issues

Error with GetSignalingPartners

Hello!!

When I try running the GetSignalingParnters using the ligand/receptor provided by the tool I get the following error:
Error in makePSOCKcluster(names = spec, ...) :
Cluster setup failed. 1 worker of 1 failed to connect.

Also, how can I construct the ligand/receptor lists for other genes?

Thanks in advance for your help.

BW

Ale

How to install package NMF 0.23.0

the PSoptSC needs the NMF>0.23.0, however the version of the NMF is only 0.22.0. Could you please tell me how to install the PSoptSC?
Thank you.

Seurat based RSoptSC

HI Mkarikom,

I wonder whether RsoptSC can applied on Seurat object?

error message when installing soptSC

Hi, I'm using R4.0.4, and when installing soptSC, I got the following message:
Error: Failed to install 'RSoptSC' from GitHub:
Failed to install 'NMF' from GitHub:
(converted from warning) installation of package ‘C:/Users/jimmy/AppData/Local/Temp/RtmpYVOa9Y/file7bc467d73d89/NMF_0.30.1.tar.gz’ had non-zero exit status

Could you help me? Thank you!
jimmy

how to download the ligand-receptor-target list

Hi,
Thanks for the wonderful tool. I may ask how to get the full ligand-receptor list?
Best,
Amare

Questions about the LR-TF database

Thank you so much for developing the tool!

I would like to know if in the LR-TF database you collected, are all LR's effects on TF activated? Is it possible that LR's effect on TF is inhibitory?

Looking forward to your reply!

Error while trying to graph cell-cell signaling circos plot

I've been trying to plot cell-cell communication circos plot with my Seurat object. The steps up to computing the matrix work fine, but I'm having trouble with SigPlot for cell-cell signaling (cluster-cluster signaling graph works just fine)

labels <- [email protected]$res.0.6
SigPlot(P = Pmats$P_agg,

•     cluster_labels = labels,
  

•     lig_cells_per_cluster = 5,
  

•     rec_cells_per_cluster = 5,
  

•     rgb_gap = 0.06,
  

•     title_text = "Signaling Between Cells")
  

Error in n + 1 : non-numeric argument to binary operator

upload 10x data

can not upload 10x reads. The 10x format gives three documents to upload: 'barcodes.tsv', 'genes.tsv', and 'matrix.mtx'. I can not upload these files without error messages. I have been using the LoadData function. Is there another way to upload the documents and create an object that is readable for the SoptSC platform in R?

Not an issue but a question

Great work!
I am wondering how easily it is to add an spatial constraint in your tool when calculating cell cell interaction probabilities?
Thanks!
Yunguan

Does this interface with Seurat?

Just curious if there is a way to integrate UMAP and cluster analysis obtained from Seurat for pseudotime analysis using this package

Errors while trying to run through vignette("Joost_et_al") demo

I've been trying to run through the demo and have been running into some bugs. First, for the logdata <- log10(data + 1), should be logdata <- log10(GSE67602_Joost$data + 1).

I'm also having trouble with the inputs for GetMarkerTable.

markers <- GetMarkerTable(counts_data = filtered_data,

•                       cluster_labels = labels,
  

•                       H = H,
  

•                       n_sorted = 25,
  

•                       gene_names = gene_names,
  

•                       gene_expression_threshold = 3,
  

•                       n_features = 20000,
  

•                       use_H = TRUE)
  

gives:
Error in GetMarkerTable(counts_data = filtered_data, cluster_labels = labels, :
unused arguments (gene_names = gene_names, n_features = 20000)

markers <- GetMarkerTable(counts_data = filtered_data,

•                       cluster_labels = labels,
  

•                       H = H,
  

•                       n_sorted = 25,
  

•                       gene_expression_threshold = 3,
  

•                       use_H = TRUE)
  

gives:
Error in matrix(rownames(counts_data), ncol = 1) :
'data' must be of a vector type, was 'NULL'

and

markers <- GetMarkerTable(counts_data = filtered_data$M_variable,

•                       cluster_labels = labels,
  

•                       H = H,
  

•                       n_sorted = 25,
  

•                       gene_expression_threshold = 3,
  

•                       use_H = TRUE)
  

gives: Error in cbind(genes, gene_assignments) :
number of rows of matrices must match (see arg 2)

So, I'm not quite sure if this is an input error or a bug.

Thanks.

Installing BiocNeighbors

Hi @mkarikom and @alevigi,

I have the same problem as in #15 (comment)_ with the "wrong number of template arguments" etc. when trying to install BiocNeighbor.

I was wondering if you could explain what you meant about editing the path and/or how to use the remote library for this?

(From the thread:

"looks like you figured it out but sometimes hpc accounts require you to edit the path via so-called module loading to enable compilation of c-based dependencies"
"I managed to finally install it using the remote library."

)

'what' must be a function or character string

Hi！
when I run the vignette clusters <- ClusterCells(similaritymatrix = S$W, n_comp = 15, .options='p'), find error
Error in do.call(getGeneric("seed"), c(list(x = x, model = init, method = seed.method), : 'what' must be a function or character string
so could anyone give some suggestion?
Thanks!!

ClusterCells Error

clusters <- ClusterCells(similarityMatrix = S$W, n_comp = 15, .options='p')
Error in (new("standardGeneric", .Data = function (x) :
unused arguments (model = list("NMFstd", 7, 0), method = "nndsvd")

I'm running the above command with your example file. Any suggestion on how to fix this error?

Error code on installing RSoptSC

I am getting the following error upon installing RSoptSC.

installing source package 'RSoptSC' ...
** using staged installation
** R
** inst
** byte-compile and prepare package for lazy loading
Error: object 'isSpike<-' is not exported by 'namespace:SingleCellExperiment'
Execution halted
ERROR: lazy loading failed for package 'RSoptSC'

Can someone please help me resolve this issue? Thank you. I have tried reinstalling SingleCellExperiment including different versions. No luck.

BiocNeighbor dependecy error

Hello, I'm trying to install RSoptsc on the cluster and I'm having problems with the BipocNeighbor dependency.
I copied and pasted the whole error message below.
I also tried to install the package separately in different ways but nothing works. I did not have this problem when I installed RSoptsc locally.

Any advice?

Thank you very much in advance.

In file included from annoy.cpp:1:0:
annoy.h:33:63: error: wrong number of template arguments (4, should be 5)
typedef AnnoyIndex<Index_t, Data_t, Distance, Kiss64Random> _index;
^
In file included from annoy.h:17:0,
from annoy.cpp:1:
/users/k1642876/R/x86_64-pc-linux-gnu-library/3.6/RcppAnnoy/include/annoylib.h:845:9: note: provided for ‘template<class S, class T, class Distance, class Random, class ThreadedBuildPolicy> class AnnoyIndex’
class AnnoyIndex : public AnnoyIndexInterface<S, T> {
^~~~~~~~~~
annoy.cpp: In constructor ‘Annoy::Annoy(int, const string&, double)’:
annoy.cpp:7:9: error: request for member ‘load’ in ‘((Annoy)this)->Annoy::obj’, which is of non-class type ‘Annoy::_index {aka int}’
obj.load(fname.c_str());
^~~~
annoy.cpp: In member function ‘MatDim_t Annoy::get_nobs() const’:
annoy.cpp:16:16: error: request for member ‘get_n_items’ in ‘((const Annoy)this)->Annoy::obj’, which is of non-class type ‘const _index {aka const int}’
return obj.get_n_items();
^~~~~~~~~~~
annoy.cpp: In member function ‘void Annoy::find_nearest_neighbors(CellIndex_t, NumNeighbors_t, bool, bool)’:
annoy.cpp:44:9: error: request for member ‘get_nns_by_item’ in ‘((Annoy)this)->Annoy::obj’, which is of non-class type ‘Annoy::_index {aka int}’
obj.get_nns_by_item(c, K + 1, get_search_k(K + 1), &kept_idx, dptr); // +1, as it forgets to discard 'self'.
^~~~~~~~~~~~~~~
annoy.cpp: In member function ‘void Annoy::find_nearest_neighbors(const double, NumNeighbors_t, bool, bool)’:
annoy.cpp:86:9: error: request for member ‘get_nns_by_vector’ in ‘((Annoy*)this)->Annoy::obj’, which is of non-class type ‘Annoy::_index {aka int}’
obj.get_nns_by_vector(holding.data(), K, get_search_k(K), &kept_idx, dptr);
^~~~~~~~~~~~~~~~~
In file included from annoy.cpp:97:0:
build_annoy.cpp: In instantiation of ‘Rcpp::RObject build_annoy_internal(Rcpp::NumericMatrix, int, const string&) [with Distance = Manhattan; Rcpp::RObject = Rcpp::RObject_ImplRcpp::PreserveStorage; Rcpp::NumericMatrix = Rcpp::Matrix<14>; std::__cxx11::string = std::__cxx11::basic_string]’:
build_annoy.cpp:26:66: required from here
build_annoy.cpp:14:13: error: request for member ‘add_item’ in ‘obj’, which is of non-class type ‘Annoy::_index {aka int}’
obj.add_item(i, tmp.data());
~~~~^~~~~~~~
build_annoy.cpp:17:9: error: request for member ‘build’ in ‘obj’, which is of non-class type ‘Annoy::_index {aka int}’
obj.build(Ntrees);
~~~~^~~~~
build_annoy.cpp:18:9: error: request for member ‘save’ in ‘obj’, which is of non-class type ‘Annoy::_index {aka int}’
obj.save(Fname.c_str());
~~~~^~~~
build_annoy.cpp: In instantiation of ‘Rcpp::RObject build_annoy_internal(Rcpp::NumericMatrix, int, const string&) [with Distance = Euclidean; Rcpp::RObject = Rcpp::RObject_ImplRcpp::PreserveStorage; Rcpp::NumericMatrix = Rcpp::Matrix<14>; std::__cxx11::string = std::__cxx11::basic_string]’:
build_annoy.cpp:28:66: required from here
build_annoy.cpp:14:13: error: request for member ‘add_item’ in ‘obj’, which is of non-class type ‘Annoy::_index {aka int}’
obj.add_item(i, tmp.data());
~~~~^~~~~~~~
build_annoy.cpp:17:9: error: request for member ‘build’ in ‘obj’, which is of non-class type ‘Annoy::_index {aka int}’
obj.build(Ntrees);
~~~~^~~~~
build_annoy.cpp:18:9: error: request for member ‘save’ in ‘obj’, which is of non-class type ‘Annoy::_index {aka int}’
obj.save(Fname.c_str());
~~~~^~~~
make: *** [annoy.o] Error 1
ERROR: compilation failed for package ‘BiocNeighbors’

• removing ‘/users/k1642876/R/x86_64-pc-linux-gnu-library/3.6/BiocNeighbors’
  Error: Failed to install 'RSoptSC' from GitHub:
  (converted from warning) installation of package ‘BiocNeighbors’ had non-zero exit status