guido.guides

Module Contents

Classes

Guide

class guido.guides.Guide(sequence, pam_position, pam_len, strand='+', max_flanking_length=75, cut_offset=3, chromosome='seq', start=0)

property location

Returns the location of the guide in the format: chr:start-end.

Returns:

str

String representation of gRNA location

property off_targets_string

Returns a string representation of the off-targets.

The string representatio captures the number of off-targets with certain number of mismatches: n0|n1|n2|n3|n4|n5 (total), where n0 is the number of off-targets with 0 mismatches, n1 is the number of off-targets with 1 mismatch, etc.

For example, if there are 3 off-targets with 0 mismatches, 2 with 1 mismatch, 1 with 2 mismatches, 0 with 3 mismatches, 5 with 4 mismatches and 1 with 5 mismatches the string representation will be “3|2|1|0|5|1 (13)”. In the parenthesis, the total number of off-targets is given.

Returns:

str

String representation of off-targets.

property layers

__repr__()

Return repr(self).

__getattr__(attr)

_create_mmej_oof_string(mmej_patterns)

simulate_end_joining(n_patterns=5, length_weight=20)

Simulate Microhomology-Mediated End Joining (MMEJ) events for the gRNA.

MMEJ scoring is based on the Bae et al. 2014 paper (https://doi.org/10.1038/nmeth.3015)

Parameters:

n_patternsint, optional

Number of top-scoring MMEJ patterns to keep, by default 5

length_weightint, optional

Lengeth weight, by default 20

find_off_targets(genome, **kwargs)

Finds off-targets for the guide. The off-targets are found using Bowtie. Bowtie index for the genome must be built before running this function.

Parameters:

genomeGenome

Genome object with the Bowtie index built

Notes

The off-targets are stored in the off_targets attribute. Based on the off-targets, the following layers are added to the guide:

• ot_sum_score: sum of the off-target scores - the lower the better

• ot_cfd_score_mean: mean of the CFD scores of the off-targets

• ot_cfd_score_max: max CFD scores of the off-targets

• ot_cfd_score_sum: sum CFD scores of the off-targets

add_layer(name, layer_data)

_summary_

Parameters:

namestr

_description_

layer_datafloat

_description_

layer(key)

_summary_

Parameters:

key_type_

_description_

Returns:

_type_

_description_

Raises:

ValueError

_description_

add_azimuth_score(model_file='V3_model_nopos.pickle')

Apply Azimuth score to a list of guides.

Azimuth is a machine learning-based predictive modelling of CRISPR/Cas9 guide efficiency. Sometimes its reffered to as Doench 2016 score.

Described in https://doi.org/10.1038/nbt.3437 (Doench et al., 2016)

Returns:

float

Azimuth score.