Regex Typo in Frangieh 2021 processing

[atong01]

Hi,

The perturbations in Frangieh 2021 processing notebook are processed using this regex line

adata.obs['perturbation_name'] = [re.sub('[_123]+', '', s) for s in adata.obs['sgRNA']]

This should be changed to '_[123]+'. This is meant to strip the guide number, right now it also strips out any 1 2 or 3 in the gene.

Unique genes in adata.obs["sgRNA"]

Index(['A2M_1', 'A2M_2', 'A2M_3', 'ACSL3_1', 'ACSL3_2', 'ACSL3_3', 'ACTA2_1',
       'ACTA2_2', 'ACTA2_3', 'AEBP1_1',
       ...
       'VDAC2_3', 'WBP2_1', 'WBP2_2', 'WBP2_3', 'WNT7A_1', 'WNT7A_2',
       'WNT7A_3', 'XAGE1A_1', 'XAGE1A_2', 'XAGE1A_3'],
      dtype='object', length=818)

In the processed data file the current line changes the first gene A2M turns into AM. This loses information and makes matching guides to the data difficult.