greenleaflab / mpal-single-cell-2019 Goto Github PK

Hi ,

I'm having some trouble with the umapProjection from
'Update scRNA_02_Cluster_Disease_w_Reference_v1.R'
at line 366:
umapProjection <- uwot::umap_transform(as.matrix(lsiProjection$matSVD)[,1:25], umapManifold, verbose = TRUE)

The error is
Error in lsiProjection$matSVD : $ operator not defined for this S4 class
but I cannot see the matSVD slot here, not even within lsiProjection@ .

As I understand, it should be coming from
lsiProjection <- projectLSI(matProjectLSI, lsiReference)

I would like to use it to project a second scRNAseq dataset on a previously determined scRNAseq umap.

Thank you for any comments and tips!

Best,
Peter

Hi @jgranja24 @jeffmgranja

I had a question regarding this sentence from the methods section:

"Matching scATAC-seq-scRNA-seq pairs using Seurat’s canonical correlation analyses.
To integrate our epigenetic and transcriptomic data we built on previous approaches for integration10,37. We found the approach that worked best for our integrative analyses was using Seurat’s CCA. We performed integration for each biological group separately because (1) it improved alignment accuracy and (2) required much less memory."

For each biological group separately, you mean for each cell type separately? Does this mean you took a multi-patient RNA dataset of only CD8T cells and aligned it with the full ATAC dataset composed of all your patients and all of your cell types?

Thanks in advance!

Can you tell me How can I download the gene.gtf file?Thanks!

Hello,
Can you tell me which the GRCh37 (hg19) reference genome you use in scRNA alignment? where to download?
I try to be consistent with you analysis.

Thank you for freely providing these datasets and analysis scripts.
I've tried to recover the gene expression and antibody features data from an original 10X bam file (sample GSM4138874) and even though I could get the the gene expression data from this file, I can't recover the antibody features reads from the raw 10X bam files.

I've tried extracting the reads using the following one liner:
samtools view original10X.bam | LC_ALL=C grep -F fb:Z | samtools view -b -o original10X_ADT.bam
The one liner is from an official 10X Genomics support page: https://kb.10xgenomics.com/hc/en-us/articles/360022448251-Is-there-way-to-filter-the-BAM-file-produced-by-10x-pipelines-with-a-list-of-barcodes-
It did not generate any output and at a read inspection I could not find reads with the fb:Z flag that is used to mark reads pertaining the antibody features.

Has anyone encountered the same issue when they tried to retrieve antibody features data from in raw 10x Bam Files from this publication? Thank you for your help!

Hi,

I understand that in the script scATAC_04_Compute_Gene_Scores.R, gene activity scores are calculated by the custom_cicero_cds function in a more efficient way and at the same time cell grouping information is kept for next steps (that is great!). It seems to me that unlike the original cicero paper method, the following codes ignores the peak-gene distance penalty but directly calculates the coaccessibility correlation and others. I wonder if I miss something or it does not matter to skip the distant penalty issue. My question: is there any difference if the peak-gene distance variable is considered? This is important to fully understand the gene activity score, so I am looking forward to hear from your suggestion.

Cheers,
Meijiao

Hi,

You are using function sparseMatTTest, but it's not defined anywhere in your script.
Could you please add the code for that part?

Several functions do not work with Seurat v3. Will the code provided be maintained for compatibility with Seurat v3?

Hi,

How did you compose those pretty scentific figures? like the panels f and g in your example figure, which show density tracks of scADT-seq and scRNA-seq. Did you plot them by customized python code or R code? I am starting to learn scentific plotting by python and thirsty for suggestions.

Dear Granja

It is quite a smart idea to apply the 'uwot_transform' to in situ projection of new data (disease sample of both single cell or bulk datasets) to a already calculated umap graph. This method, which is named 'LSI_projection' is exemplified in your script scRNA_02_Cluster_Disease_w_Reference_v1.R. However, when loading the saved umap model file from a tar file in 'scRNA-Projection-UMAP.zip', it complained for the unloaded nn_index. I looked into this issue further by run the load_uwot() code step by step and found a error message "Index size is not a multiple of vector size: Success (0)". and if applied model$nn_index$getNTrees() , I got zero. But , the nn1 file is not empty. As I described above, it seems that either the nn_index is empty or is not compatible with my R enviroment.

Help!

Meijiao

Hi there,

I got a problem that when I run scRNA_01, there is an error says no seuratSNN function. Could you please check it ?

Thanks!

Hi,
I've tried to run scRNA_01_Clustering_UMAP_v1.R using default parameters, and the
"scRNA-All-Hematopoiesis-MPAL-191120.rds" as input. When I do run the
#Optimize LSI Features
lsiObj <- optimizeLSI(assay(se),
resolution = resolution,
pcsUse = nPCs,
varFeatures = nTop)

it gives error after LSI 1 finished
Number of communities: 38
Elapsed time: 188 seconds
Additional LSI 2...
Error in if (n < 0L) max(length(x) + n, 0L) else min(n, length(x)) :
missing value where TRUE/FALSE needed

Anyone has faced with this error?

I am trying to run scRNA_01_Clustering_UMAP_v1.R, and apparently the function optimizeLSI is iterative. The code shows that each iteration take 30 seconds and the maximum number of iterations is 3000. Am I doing something wrong or is it supposed to be this slow?

Hi, can you share me the markers you used for scRNA-seq cell cell type annotation.Thanks.

Hi,

Many thanks for the repository of code and data. I was hoping to use some of the healthy scATAC datasets, but after downloading the healthy hematopoiesis scATAC cell x peak SummarizedExperiment object, I found that the cell names (column names) are all in the form BMM_R1_15, BMM_R1_16... etc.

However, the fragment files when I download them have the 10X cell barcodes. Is there a way I can match the barcodes back to the processed SummarizedExperiment objects? I've looked in the various metadata slots for the SummarizedExperiment object but can't seem to find anything.

Many thanks in advance.

Can you tell me How can I download the bam file for scATAC-seq data?
Thanks!

Hi there,

Just wondering if there is a scATAC peak file associated with the published data set that could be used for reference?

Thanks!

Making Seurat Object...
Error in CreateAssayObject(counts = counts, min.cells = min.cells, min.features = min.features) :
No feature names (rownames) names present in the input matrix

Dear authors,
In the README, the download link for scRNA-Seq data of Healthy Hematopoiesis, which has scATAC in its name, does not match with its description, scRNA-seq. Could you please check it ?

scRNA-seq Hematopoeisis cell x gene Summarized Experiment :
https://jeffgranja.s3.amazonaws.com/MPAL-10x/Supplementary_Data/Healthy-Data/scATAC-Cicero-GA-Hematopoiesis-191120.rds

Many thanks

scADT-All-Hematopoiesis-MPAL-191120.rds contains the rawCounts but not the CLR normalised assay.
I tried applying compositions::clr as well as the CLR transformation implemented in Seurat :

clr_function <- function(x) { 
 return(log1p(x = x / (exp(x = sum(log1p(x = x[x > 0]), na.rm = TRUE) / length(x = x)))))
}

but I cannot reproduce the normalisde values in the CLR assay from scADT-Healthy-Hematopoiesis-191120.rds. Could you add the code that you used for the CLR normalisation?

Thank you,
Ricard.

Thank you very much for providing so much data and code in such an organized fashion.
I am interested in the LSI projections and reclassification of the AML cells as in fig 7.
Perhaps it should be in the file 'scRNA-AML-Analyses.zip', however, it seems that the files in the folder '__MACOSX' are corrupted.
best,
Eldad

I went through your pipeline and the part that is producing the first two input files for scRNA_scATAC_Integration_04_Link_TFs_To_Targets.R script is missing. Could you please clarify about this part futher?

diff_scATAC_file <- "Save-scATAC-PromoterNorm-Differentials.rds"
diff_scRNA_file <- "output/scATAC_scRNA/Differential/Differential_Summary.rds"

keepFilteredChromosomes function at line 310 of scATAC_04_Compute_Gene_Scores.R seems invalid and can not be found in any R packages (for example GenomicRanges, IRanges). Did you defined this function as your own and forget to upload?

Thanks

When I tried running the scripts, I got the following issue. There is no definition for the function "keepFilteredChromosomes". Could you please tell me where to find this function?