seurat_obj.rds not found while running downstream analysis

[khushi0810]

Hello Wenbao @wbaopaul ,

I am using v1.5.1 on one of my samples , and I'm using scATAC-pro reprocess because I have my bam files from cell ranger and then using scATAC-pro downstream .

I am running on HPC as a batch job. I am attaching my log file , configure file as well as my script for you reference.

Any suggestions are highly appreciated.

Thank you
Khushi
reprocess_cellranger_output.log
configure_user.txt

#####################################

module load singularity/3.9.9

cd /project/bioinformatics/JLee_lab/s227606/scATAC/NONP

singularity exec --bind /project/bioinformatics/JLee_lab/s227606/scATAC -H /project/bioinformatics/JLee_lab/s227606/scATAC/NONP --cleanenv /project/bioinformatics/JLee_lab/s227606/scATAC/NONP/scatac-pro_latest.sif scATAC-pro -s reprocess_cellranger_output -i /project/bioinformatics/JLee_lab/s227606/scATAC/files/BL51_NP/outs/possorted_bam.bam,/project/bioinformatics/JLee_lab/s227606/scATAC/files/BL51_NP/outs/fragments.tsv.gz -c /project/bioinformatics/JLee_lab/s227606/scATAC/configure_user.txt

singularity exec --bind /project/bioinformatics/JLee_lab/s227606/scATAC -H /project/bioinformatics/JLee_lab/s227606/scATAC/NONP --cleanenv /project/bioinformatics/JLee_lab/s227606/scATAC/NONP/scatac-pro_latest.sif scATAC-pro -s downstream -i /project/bioinformatics/JLee_lab/s227606/scATAC/NONP/output/filtered_matrix/MACS2/FILTER/matrix.mtx -c /project/bioinformatics/JLee_lab/s227606/scATAC/configure_user.txt

###################################################

[wbaopaul]

Can you check some intermediate results? Such as how many cells did you get under filtered_Matrix folder or any other output? I guess some of your filtering parameters may be too stringent, like min_frac_peak set as 0.5.

[khushi0810]

Hello,

Thank you so much for your reply. I am attaching my results under filtered_matrix folder.
Should I try increasing the threshhold ?

Thanks
barcodes.txt
features.txt

[wbaopaul]

No, you need to decrease your thresholds, for instance, min_uniq_frags to 2000, min_frac_peak 0.25. You can increase max_frac_mito to 0.15 or 0.2.

[khushi0810]

Thank you so much. I changed the threshholds and was able to resolve this error but bumped into another error

Error in FindNeighbors.Seurat(seurat.obj, reduction = "pca", dims = 1:30, :
More dimensions specified in dims than have been computed
Calls: FindNeighbors -> FindNeighbors.Seurat
Execution halted

Should I change the nReduction parameter ?

Error: Cannot find 'active_clusters' in this Seurat object
Execution halted
Report generation Done!

Do I need to change this parameter ?
K_CLUSTERS = 0.2 ## the number of cluster (in integer) or the resolution parameter (in float) for louvain algorithm (implemented by seurat)

Thank you !

[wbaopaul]

Not sure what exactly happened. Have you checked the outputs? I guess you probably ended up too few number of cells.