Pipeline philosophy

Pipeline philosophy:

Single-cell data analysis cannot be done through a one-shot pipeline. Indeed, the parameters used in the pipeline strongly depend on the quality of the data and the biological variability (due to the type of tissue studied). So, you have to estimate them based on the previous steps. For example, for the filter thresholds of poor quality droplets or empty droplets, the quality of the droplets in the sample should be studied.

Thus, there are many interruptions in the pipeline that must be performed, materialized by steps:

• Alignment_countTable_GE,
• Droplets_QC_GE,
• Filtering_GE,
• Norm_DimRed_Eval_GE,
• Clust_Markers_Annot_GE,
• Cerebro,
• Alignment_countTable_ADT,
• Adding_ADT,
• Alignment_annotations_TCR_BCR,
• Adding_TCR,
• Adding_BCR.

This step system also makes it easy to rerun specific steps to test multiple parameters and compare results, without having to rerun the entire pipeline.

Also, it allows to choose the type of analysis to be carried out according to the type of data (gene expression, cell surface proteins, immune repertoire profiling, etc.), and thus to be able to carry out multiomics analysis, all combined in a single pipeline of analysis.

Names of samples:

To help distinguish between different types of data easier, a tag is added as a suffix to the name of the samples:

• "_GE": for Gene Expressions
• "_ADT": for Antibody-Derived Tags (Cell Surface Proteins corresponding to the CITE-seq part)
• "_TCR": for T Cell Receptor expression
• "_BCR": for B Cell Receptor expression

If the tag is already at the end of your sample, it will not be added.