Input Data

tribal requires the input data to be first clustered into clonotypes, i.e., groups of cells that descend from the same naive B cell receptor. We recommend Dandelion to assist with this preprocessing step.

tribal requires two input files for preprocessing. The single-cell RNA sequencing data and the germline root sequences.

Sequencing data

Sequencing data should be provided in a csv file with the following columns:

Column Name	Type	Description	Required
cell	str or int	unqiue id or barcode of the sequnced B cell	True
clonotype	str	unique clonotype id to which that cell belongs	True
heavy_chain_isotype	str	the isotype of the constant region of the heavy chain	True
heavy_chain_seq	str	the variable region sequence of the heavy chain	True
heavy_chain_allele	str	the v allele of the heavy chain	True
heavy_chain_isotype	str	the isotype of the constant region of the heavy chain	True
light_chain_seq	str	the variable region sequence of the light chain	False
light_chain_allele	str	the v allele of the light chain	False, if light_chain_seq not provided

See data.csv for an example.

Germline clonotype roots

Additionally, the germline root sequences by clonotype should be provided in a csv file containing the heavy chain sequence and optionally, the light chain sequence.

Column Name	Type	Description	Required
clonotype	str	unique clonotype id of the germline root (naive BCR)	True
heavy_chain_root	str	the heavy chain variable region germline root sequence	True
light_chain_root	str	the light chain variable region germline root sequence	False

See roots.csv for an example.

Note

All light chain columns may be omitted if the use_light_chain argument in preprocess is False. In other words, tribal may be used with only the heavy chain BCR sequences.