Before using the pipeline the following files need to be provided/adapted:
- config.yaml
- samplemap
- HashingFile
- featureReferenceFile
Before running the pipeline the config.yaml
file needs to be adapted to contain the input and output paths for the intended analysis.
Adaptation necessary for a default run:
- In section [
inputOutput
] the input directoriesinput_fastqs_gex
,input_fastqs_adt
need to point to the location of the respective FASTQ files. This location needs to be specified relative to the gExcite working directory (usuallygExcite_pipeline/
). - In section [
resources
],reference_transcriptome
needs to point to the location of the genomic reference used for the cellranger mapping - In sections [
tools
][cellranger_count_gex
] and [tools
][cellranger_count_adt
],call
needs to point to the the path to the cellranger installation - Section [
computingResources
] needs to list the resources that can be assigned to the analysis steps your data and batch system. The example resources specify memory per job, not thread. - Section [
scampi
][resources
] needs to be filled with the cell type information, selected genes to show in expression plots, and gene sets for the GSVA analysis.
The default example config file is pointing to the example data input files.
Further, a "samplemap" must be provided specifying sample-specific parameters in a tab-delimited text file. A pre-configured samplemap ready to run on the test data that can be adapted is provided in this directory.
Example samplemap:
sample HashingFile nTargetCells featureReferenceFile
sampleA HashingFileA 10000 featureReferenceFileA
sampleB HashingFileB 15000 featureReferenceFileB
With one line per set of samples
-
sample
contains the sample identifier that is used throughout the pipeline -
HashingFile
contains the full path to the comma separated text file containing the hashtag barcodes and their assignment to individual sample names (see HashingFile). -
nTargetCells
corresponds to the number of targeted cells for the sample. -
featureReferenceFile
corresponds to the ADT feature reference file for the sample set. For further information please consult the Cellranger tool documentation.
NOTE: the path to the FeatureReferenceFile must be relative to the gExcite working directory (usuallygExcite_pipeline/
).
In case of hashed samples, the hashtag barcodes, the hashtag names, and the corresponding sample names must be associated with the sample set. To do so, we need a comma-separated file with the following structure:
Barcode1,TagName1,sampleA
Barcode2,TagName2,sampleB
The "featureReferenceFile" is a comma-separated text file describing all ADT antibodies used in the experiment at hand.
For further information please consult the Cellranger tool documentation.
An example feature_reference.txt
ready to run on the test data is available in the testdata
directory.