rs3 0.0.15

Predict the activity of CRISPR sgRNAs

Rule Set 3

Python package to predict the activity of CRISPR sgRNA sequences using Rule Set 3

Install

You can install the latest release of rs3 from pypi using

pip install rs3

For mac users you may also have to brew install the OpenMP library

brew install libomp

or install lightgbm without Openmp

pip install lightgbm --install-option=--nomp

See the LightGBM documentation for more information

Documentation

You can see the complete documentation for Rule Set 3 here.

Quick Start

Sequence based model

To calculate Rule Set 3 (sequence) scores, import the predict_seq function from the seq module.

from rs3.seq import predict_seq

You can store the 30mer context sequences you want to predict as a list.

context_seqs = ['GACGAAAGCGACAACGCGTTCATCCGGGCA', 'AGAAAACACTAGCATCCCCACCCGCGGACT']

You can specify the Hsu2013 or Chen2013 as the tracrRNA to score with. We generally find any tracrRNA that does not have a T in the fifth position is better predicted with the Chen2013 input.

predict_seq(context_seqs, sequence_tracr='Hsu2013')

Calculating sequence-based features


100%|██████████| 2/2 [00:00<00:00, 15.04it/s]





array([-0.90030944,  1.11451622])

Target based model

To get target scores, which use features at the endogenous target site to make predictions, you must build or load feature matrices for the amino acid sequences, conservation scores, and protein domains.

As an example, we'll calculate target scores for 250 sgRNAs in the GeckoV2 library.

import pandas as pd
from rs3.predicttarg import predict_target
from rs3.targetfeat import (add_target_columns,
                            get_aa_subseq_df,
                            get_protein_domain_features,
                            get_conservation_features)

design_df = pd.read_table('test_data/sgrna-designs.txt')
design_df.head()

<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}

</style>

	Input	Quota	Target Taxon	Target Gene ID	Target Gene Symbol	Target Transcript	Target Reference Coords	Target Alias	CRISPR Mechanism	Target Domain	...	On-Target Rank Weight	Off-Target Rank Weight	Combined Rank	Preselected As	Matching Active Arrayed Oligos	Matching Arrayed Constructs	Pools Containing Matching Construct	Pick Order	Picking Round	Picking Notes
0	PSMB7	2	9606	ENSG00000136930	PSMB7	ENST00000259457.8	NaN	NaN	CRISPRko	CDS	...	1.0	1.0	7	GCAGATACAAGAGCAACTGA	NaN	BRDN0004619103	NaN	1	0	Preselected
1	PSMB7	2	9606	ENSG00000136930	PSMB7	ENST00000259457.8	NaN	NaN	CRISPRko	CDS	...	1.0	1.0	48	AAAACTGGCACGACCATCGC	NaN	NaN	NaN	2	0	Preselected
2	PRC1	2	9606	ENSG00000198901	PRC1	ENST00000394249.8	NaN	NaN	CRISPRko	CDS	...	1.0	1.0	7	AAAAGATTTGCGCACCCAAG	NaN	NaN	NaN	1	0	Preselected
3	PRC1	2	9606	ENSG00000198901	PRC1	ENST00000394249.8	NaN	NaN	CRISPRko	CDS	...	1.0	1.0	8	CTTTGACCCAGACATAATGG	NaN	NaN	NaN	2	0	Preselected
4	TOP1	2	9606	ENSG00000198900	TOP1	ENST00000361337.3	NaN	NaN	CRISPRko	CDS	...	1.0	1.0	1	NaN	NaN	BRDN0001486452	NaN	2	1	NaN

5 rows × 60 columns

Throughout the analysis we will be using a core set of ID columns to merge the feature matrices. These ID columns should uniquely identify an sgRNA and its target site.

id_cols = ['sgRNA Context Sequence', 'Target Cut Length', 'Target Transcript', 'Orientation']

Amino acid sequence input

To calculate the amino acid sequence matrix, you must first load the complete sequence from ensembl using the build_transcript_aa_seq_df. See the documentation for the predicttarg module for an example of how to use this function.

In this example we will use amino acid sequences that have been precalculated using the write_transcript_data function in the targetdata module. Check out the documentation for this module for more information on how to use this function.

We use pyarrow to read the written transcript data. The stored transcripts are indexed by their Ensembl ID without the version number identifier. To get this shortened version of the Ensembl ID use the add_target_columns function from the targetfeat module. This function adds the 'Transcript Base' column as well as a column indicating the amino acid index ('AA Index') of the cut site. The 'AA Index' column will be used for merging with the amino acid translations.

design_targ_df = add_target_columns(design_df)
design_targ_df.head()

<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}

</style>

	Input	Quota	Target Taxon	Target Gene ID	Target Gene Symbol	Target Transcript	Target Reference Coords	Target Alias	CRISPR Mechanism	Target Domain	...	Combined Rank	Preselected As	Matching Active Arrayed Oligos	Matching Arrayed Constructs	Pools Containing Matching Construct	Pick Order	Picking Round	Picking Notes	AA Index	Transcript Base
0	PSMB7	2	9606	ENSG00000136930	PSMB7	ENST00000259457.8	NaN	NaN	CRISPRko	CDS	...	7	GCAGATACAAGAGCAACTGA	NaN	BRDN0004619103	NaN	1	0	Preselected	64	ENST00000259457
1	PSMB7	2	9606	ENSG00000136930	PSMB7	ENST00000259457.8	NaN	NaN	CRISPRko	CDS	...	48	AAAACTGGCACGACCATCGC	NaN	NaN	NaN	2	0	Preselected	46	ENST00000259457
2	PRC1	2	9606	ENSG00000198901	PRC1	ENST00000394249.8	NaN	NaN	CRISPRko	CDS	...	7	AAAAGATTTGCGCACCCAAG	NaN	NaN	NaN	1	0	Preselected	106	ENST00000394249
3	PRC1	2	9606	ENSG00000198901	PRC1	ENST00000394249.8	NaN	NaN	CRISPRko	CDS	...	8	CTTTGACCCAGACATAATGG	NaN	NaN	NaN	2	0	Preselected	263	ENST00000394249
4	TOP1	2	9606	ENSG00000198900	TOP1	ENST00000361337.3	NaN	NaN	CRISPRko	CDS	...	1	NaN	NaN	BRDN0001486452	NaN	2	1	NaN	140	ENST00000361337

5 rows × 62 columns

transcript_bases = design_targ_df['Transcript Base'].unique()
transcript_bases[0:5]

array(['ENST00000259457', 'ENST00000394249', 'ENST00000361337',
       'ENST00000368328', 'ENST00000610426'], dtype=object)

aa_seq_df = pd.read_parquet('test_data/target_data/aa_seqs.pq', engine='pyarrow',
                            filters=[[('Transcript Base', 'in', transcript_bases)]])
aa_seq_df.head()

<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}

</style>

	Target Transcript	Target Total Length	Transcript Base	version	seq	molecule	desc	id	AA len
0	ENST00000259457.8	834	ENST00000259457	3	MAAVSVYAPPVGGFSFDNCRRNAVLEADFAKRGYKLPKVRKTGTTI...	protein	None	ENSP00000259457	277
1	ENST00000394249.8	1863	ENST00000394249	3	MRRSEVLAEESIVCLQKALNHLREIWELIGIPEDQRLQRTEVVKKH...	protein	None	ENSP00000377793	620
2	ENST00000361337.3	2298	ENST00000361337	2	MSGDHLHNDSQIEADFRLNDSHKHKDKHKDREHRHKEHKKEKDREK...	protein	None	ENSP00000354522	765
3	ENST00000368328.5	267	ENST00000368328	4	MALSTIVSQRKQIKRKAPRGFLKRVFKRKKPQLRLEKSGDLLVHLN...	protein	None	ENSP00000357311	88
4	ENST00000610426.5	783	ENST00000610426	1	MPQNEYIELHRKRYGYRLDYHEKKRKKESREAHERSKKAKKMIGLK...	protein	None	ENSP00000483484	260

From the complete transcript translations, we extract an amino acid subsequence as input to our model. The subsequence is centered around the amino acid encoded by the nucleotide preceding the cut site in the direction of transcription. This is the nucleotide that corresponds to the 'Target Cut Length' in a CRISPick design file. We take 16 amino acids on either side of the cut site for a total sequence length of 33.

The get_aa_subseq_df from the targetfeat module will calculate these subsequences from the complete amino acid sequences.

aa_subseq_df = get_aa_subseq_df(sg_designs=design_targ_df, aa_seq_df=aa_seq_df, width=16,
                                id_cols=id_cols)
aa_subseq_df.head()

<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}

</style>

	Target Transcript	Target Total Length	Transcript Base	version	seq	molecule	desc	id	AA len	Target Cut Length	Orientation	sgRNA Context Sequence	AA Index	extended_seq	AA 0-Indexed	AA 0-Indexed padded	seq_start	seq_end	AA Subsequence
0	ENST00000259457.8	834	ENST00000259457	3	MAAVSVYAPPVGGFSFDNCRRNAVLEADFAKRGYKLPKVRKTGTTI...	protein	None	ENSP00000259457	277	191	sense	TGGAGCAGATACAAGAGCAACTGAAGGGAT	64	-----------------MAAVSVYAPPVGGFSFDNCRRNAVLEADF...	63	80	64	96	GVVYKDGIVLGADTRATEGMVVADKNCSKIHFI
1	ENST00000259457.8	834	ENST00000259457	3	MAAVSVYAPPVGGFSFDNCRRNAVLEADFAKRGYKLPKVRKTGTTI...	protein	None	ENSP00000259457	277	137	sense	CCGGAAAACTGGCACGACCATCGCTGGGGT	46	-----------------MAAVSVYAPPVGGFSFDNCRRNAVLEADF...	45	62	46	78	AKRGYKLPKVRKTGTTIAGVVYKDGIVLGADTR
2	ENST00000394249.8	1863	ENST00000394249	3	MRRSEVLAEESIVCLQKALNHLREIWELIGIPEDQRLQRTEVVKKH...	protein	None	ENSP00000377793	620	316	sense	TAGAAAAAGATTTGCGCACCCAAGTGGAAT	106	-----------------MRRSEVLAEESIVCLQKALNHLREIWELI...	105	122	106	138	EEGETTILQLEKDLRTQVELMRKQKKERKQELK
3	ENST00000394249.8	1863	ENST00000394249	3	MRRSEVLAEESIVCLQKALNHLREIWELIGIPEDQRLQRTEVVKKH...	protein	None	ENSP00000377793	620	787	antisense	TGGCCTTTGACCCAGACATAATGGTGGCCA	263	-----------------MRRSEVLAEESIVCLQKALNHLREIWELI...	262	279	263	295	WDRLQIPEEEREAVATIMSGSKAKVRKALQLEV
4	ENST00000361337.3	2298	ENST00000361337	2	MSGDHLHNDSQIEADFRLNDSHKHKDKHKDREHRHKEHKKEKDREK...	protein	None	ENSP00000354522	765	420	antisense	AAATACTCACTCATCCTCATCTCGAGGTCT	140	-----------------MSGDHLHNDSQIEADFRLNDSHKHKDKHK...	139	156	140	172	GYFVPPKEDIKPLKRPRDEDDADYKPKKIKTED

Lite Scores

You now have all the information you need to calculate "lite" Target Scores, which are less data intensive than complete target scores, with the predict_target function from the predicttarg module.

lite_predictions = predict_target(design_df=design_df,
                                  aa_subseq_df=aa_subseq_df)
design_df['Target Score Lite'] = lite_predictions
design_df.head()

/opt/anaconda3/envs/rs3/lib/python3.8/site-packages/sklearn/base.py:310: UserWarning: Trying to unpickle estimator SimpleImputer from version 1.0.dev0 when using version 0.24.2. This might lead to breaking code or invalid results. Use at your own risk.
  warnings.warn(
/opt/anaconda3/envs/rs3/lib/python3.8/site-packages/sklearn/base.py:310: UserWarning: Trying to unpickle estimator Pipeline from version 1.0.dev0 when using version 0.24.2. This might lead to breaking code or invalid results. Use at your own risk.
  warnings.warn(

<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}

</style>

	Input	Quota	Target Taxon	Target Gene ID	Target Gene Symbol	Target Transcript	Target Reference Coords	Target Alias	CRISPR Mechanism	Target Domain	...	Off-Target Rank Weight	Combined Rank	Preselected As	Matching Active Arrayed Oligos	Matching Arrayed Constructs	Pools Containing Matching Construct	Pick Order	Picking Round	Picking Notes	Target Score Lite
0	PSMB7	2	9606	ENSG00000136930	PSMB7	ENST00000259457.8	NaN	NaN	CRISPRko	CDS	...	1.0	7	GCAGATACAAGAGCAACTGA	NaN	BRDN0004619103	NaN	1	0	Preselected	0.012467
1	PSMB7	2	9606	ENSG00000136930	PSMB7	ENST00000259457.8	NaN	NaN	CRISPRko	CDS	...	1.0	48	AAAACTGGCACGACCATCGC	NaN	NaN	NaN	2	0	Preselected	0.048338
2	PRC1	2	9606	ENSG00000198901	PRC1	ENST00000394249.8	NaN	NaN	CRISPRko	CDS	...	1.0	7	AAAAGATTTGCGCACCCAAG	NaN	NaN	NaN	1	0	Preselected	-0.129234
3	PRC1	2	9606	ENSG00000198901	PRC1	ENST00000394249.8	NaN	NaN	CRISPRko	CDS	...	1.0	8	CTTTGACCCAGACATAATGG	NaN	NaN	NaN	2	0	Preselected	0.061647
4	TOP1	2	9606	ENSG00000198900	TOP1	ENST00000361337.3	NaN	NaN	CRISPRko	CDS	...	1.0	1	NaN	NaN	BRDN0001486452	NaN	2	1	NaN	-0.009100

5 rows × 61 columns

If you would like to calculate full target scores then follow the sections below.

Protein domain input

To calculate full target scores you will also need inputs for protein domains and conservation.

The protein domain input should have 16 binary columns for 16 different protein domain sources in addition to the id_cols. The protein domain sources are 'Pfam', 'PANTHER', 'HAMAP', 'SuperFamily', 'TIGRfam', 'ncoils', 'Gene3D', 'Prosite_patterns', 'Seg', 'SignalP', 'TMHMM', 'MobiDBLite', 'PIRSF', 'PRINTS', 'Smart', 'Prosite_profiles'. These columns should be kept in order when inputting for scoring.

In this example we will load the protein domain information from a parquet file, which was written using write_transcript_data function in the targetdata module. You can also query transcript data on the fly, by using the build_translation_overlap_df function. See the documentation for the predicttarg module for more information on how to do this.

domain_df = pd.read_parquet('test_data/target_data/protein_domains.pq', engine='pyarrow',
                            filters=[[('Transcript Base', 'in', transcript_bases)]])
domain_df.head()

<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}

</style>

	type	cigar_string	id	hit_end	feature_type	description	seq_region_name	end	hit_start	translation_id	interpro	hseqname	Transcript Base	align_type	start
0	Pfam		PF12465	36	protein_feature	Proteasome beta subunit, C-terminal	ENSP00000259457	271	1	976188	IPR024689	PF12465	ENST00000259457	None	235
1	Pfam		PF00227	190	protein_feature	Proteasome, subunit alpha/beta	ENSP00000259457	221	2	976188	IPR001353	PF00227	ENST00000259457	None	41
2	PRINTS		PR00141	0	protein_feature	Peptidase T1A, proteasome beta-subunit	ENSP00000259457	66	0	976188	IPR000243	PR00141	ENST00000259457	None	51
3	PRINTS		PR00141	0	protein_feature	Peptidase T1A, proteasome beta-subunit	ENSP00000259457	182	0	976188	IPR000243	PR00141	ENST00000259457	None	171
4	PRINTS		PR00141	0	protein_feature	Peptidase T1A, proteasome beta-subunit	ENSP00000259457	193	0	976188	IPR000243	PR00141	ENST00000259457	None	182

Now to transform the domain_df into a wide form for model input, we use the get_protein_domain_features function from the targetfeat module.

domain_feature_df = get_protein_domain_features(design_targ_df, domain_df, id_cols=id_cols)
domain_feature_df.head()

<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}

</style>

	sgRNA Context Sequence	Target Cut Length	Target Transcript	Orientation	Pfam	PANTHER	HAMAP	SuperFamily	TIGRfam	ncoils	Gene3D	Prosite_patterns	Seg	SignalP	TMHMM	MobiDBLite	PIRSF	PRINTS	Smart	Prosite_profiles
0	AAAAGAATGATGAAAAGACACCACAGGGAG	244	ENST00000610426.5	sense	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0
1	AAAAGAGCCATGAATCTAAACATCAGGAAT	640	ENST00000223073.6	sense	0	1	0	0	0	1	0	0	1	0	0	1	0	0	0	0
2	AAAAGCGCCAAATGGCCCGAGAATTGGGAG	709	ENST00000331923.9	sense	0	1	0	0	0	0	0	0	0	0	0	1	0	0	0	0
3	AAACAGAAAAAGTTAAAATCACCAAGGTGT	496	ENST00000283882.4	sense	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0
4	AAACAGATGGAAGATGCTTACCGGGGGACC	132	ENST00000393047.8	sense	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0

For input into the predict_target function, the domain_feature_df should have the id_cols as well as columns for each of the 16 protein domain features.

Conservation input

Finally, for the full target model you need to calculate conservation features. The conservation features represent conservation across evolutionary time at the sgRNA cut site and are quantified using PhyloP scores. These scores are available for download by the UCSC genome browser for hg38 (phyloP100way), and mm39 (phyloP35way).

Within this package we query conservation scores using the UCSC genome browser's REST API. To get conservation scores, you can use the build_conservation_df function from the targetdata module. Here we load conservation scores, which were written to parquet using the write_conservation_data function from the targetdata module.

conservation_df = pd.read_parquet('test_data/target_data/conservation.pq', engine='pyarrow',
                                  filters=[[('Transcript Base', 'in', transcript_bases)]])
conservation_df.head()

<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}

</style>

	exon_id	genomic position	conservation	Transcript Base	target position	chromosome	genome	translation length	Target Transcript	Strand of Target	Target Total Length	ranked_conservation
0	ENSE00001866322	124415425.0	6.46189	ENST00000259457	1	9	hg38	277	ENST00000259457.8	-	834	0.639089
1	ENSE00001866322	124415424.0	7.48071	ENST00000259457	2	9	hg38	277	ENST00000259457.8	-	834	0.686451
2	ENSE00001866322	124415423.0	6.36001	ENST00000259457	3	9	hg38	277	ENST00000259457.8	-	834	0.622902
3	ENSE00001866322	124415422.0	6.36001	ENST00000259457	4	9	hg38	277	ENST00000259457.8	-	834	0.622902
4	ENSE00001866322	124415421.0	8.09200	ENST00000259457	5	9	hg38	277	ENST00000259457.8	-	834	0.870504

We normalize conservation scores to a within-gene percent rank, in the 'ranked_conservation' column, in order to make scores comparable across genes and genomes. Note that a rank of 0 indicates the least conserved nucleotide and a rank of 1 indicates the most conserved.

To featurize the conservation scores, we average across a window of 4 and 32 nucleotides centered around the nucleotide preceding the cut site in the direction of transcription. Note that this nucleotide is the 2nd nucleotide in the window of 4 and the 16th nucleotide in the window of 32.

We use the get_conservation_features function from the targetfeat module to get these features from the conservation_df.

For the predict_targ function, we need the id_cols and the columns 'cons_4' and 'cons_32' in the conservation_feature_df.

conservation_feature_df = get_conservation_features(design_targ_df, conservation_df,
                                                    small_width=2, large_width=16,
                                                    conservation_column='ranked_conservation',
                                                    id_cols=id_cols)
conservation_feature_df

<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}

</style>

	sgRNA Context Sequence	Target Cut Length	Target Transcript	Orientation	cons_4	cons_32
0	AAAAGAATGATGAAAAGACACCACAGGGAG	244	ENST00000610426.5	sense	0.218231	0.408844
1	AAAAGAGCCATGAATCTAAACATCAGGAAT	640	ENST00000223073.6	sense	0.129825	0.278180
2	AAAAGCGCCAAATGGCCCGAGAATTGGGAG	709	ENST00000331923.9	sense	0.470906	0.532305
3	AAACAGAAAAAGTTAAAATCACCAAGGTGT	496	ENST00000283882.4	sense	0.580556	0.602708
4	AAACAGATGGAAGATGCTTACCGGGGGACC	132	ENST00000393047.8	sense	0.283447	0.414293
...	...	...	...	...	...	...
395	TTTGATTGCATTAAGGTTGGACTCTGGATT	246	ENST00000249269.9	sense	0.580612	0.618707
396	TTTGCCCACAGCTCCAAAGCATCGCGGAGA	130	ENST00000227618.8	sense	0.323770	0.416368
397	TTTTACAGTGCGATGTATGATGTATGGCTT	119	ENST00000338366.6	sense	0.788000	0.537417
398	TTTTGGATCTCGTAGTGATTCAAGAGGGAA	233	ENST00000629496.3	sense	0.239630	0.347615
399	TTTTTGTTACTACAGGTTCGCTGCTGGGAA	201	ENST00000395840.6	sense	0.693767	0.639044

400 rows × 6 columns

Full Target Scores

In order to calculate Target Scores you must input the feature matrices and design_df to the predict_target function from the predicttarg module.

target_predictions = predict_target(design_df=design_df,
                                    aa_subseq_df=aa_subseq_df,
                                    domain_feature_df=domain_feature_df,
                                    conservation_feature_df=conservation_feature_df,
                                    id_cols=id_cols)
design_df['Target Score'] = target_predictions
design_df.head()

/opt/anaconda3/envs/rs3/lib/python3.8/site-packages/sklearn/base.py:310: UserWarning: Trying to unpickle estimator SimpleImputer from version 1.0.dev0 when using version 0.24.2. This might lead to breaking code or invalid results. Use at your own risk.
  warnings.warn(
/opt/anaconda3/envs/rs3/lib/python3.8/site-packages/sklearn/base.py:310: UserWarning: Trying to unpickle estimator Pipeline from version 1.0.dev0 when using version 0.24.2. This might lead to breaking code or invalid results. Use at your own risk.
  warnings.warn(

<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}

</style>

	Input	Quota	Target Taxon	Target Gene ID	Target Gene Symbol	Target Transcript	Target Reference Coords	Target Alias	CRISPR Mechanism	Target Domain	...	Combined Rank	Preselected As	Matching Active Arrayed Oligos	Matching Arrayed Constructs	Pools Containing Matching Construct	Pick Order	Picking Round	Picking Notes	Target Score Lite	Target Score
0	PSMB7	2	9606	ENSG00000136930	PSMB7	ENST00000259457.8	NaN	NaN	CRISPRko	CDS	...	7	GCAGATACAAGAGCAACTGA	NaN	BRDN0004619103	NaN	1	0	Preselected	0.012467	0.152037
1	PSMB7	2	9606	ENSG00000136930	PSMB7	ENST00000259457.8	NaN	NaN	CRISPRko	CDS	...	48	AAAACTGGCACGACCATCGC	NaN	NaN	NaN	2	0	Preselected	0.048338	0.064880
2	PRC1	2	9606	ENSG00000198901	PRC1	ENST00000394249.8	NaN	NaN	CRISPRko	CDS	...	7	AAAAGATTTGCGCACCCAAG	NaN	NaN	NaN	1	0	Preselected	-0.129234	-0.063012
3	PRC1	2	9606	ENSG00000198901	PRC1	ENST00000394249.8	NaN	NaN	CRISPRko	CDS	...	8	CTTTGACCCAGACATAATGG	NaN	NaN	NaN	2	0	Preselected	0.061647	-0.126357
4	TOP1	2	9606	ENSG00000198900	TOP1	ENST00000361337.3	NaN	NaN	CRISPRko	CDS	...	1	NaN	NaN	BRDN0001486452	NaN	2	1	NaN	-0.009100	-0.234410

5 rows × 62 columns

Target Scores can be added directly to the sequence scores for your final Rule Set 3 predictions.

Predict Function

If you don't want to generate the target matrices themselves, you can use the predict function from the predict module.

from rs3.predict import predict
import matplotlib.pyplot as plt
import gpplot
import seaborn as sns

Preloaded data

In this first example with the predict function, we calculate predictions for GeckoV2 sgRNAs.

In this example the amino acid sequences, protein domains and conservation scores were prequeried using the write_transcript_data and write_conservation_data functions from the targetdata module. Pre-querying these data can be helpful for large scale design runs.

You can also use the predict function without pre-querying and calculate scores on the fly. You can see an example of this in the next section.

The predict function allows for parallel computation for querying databases (n_jobs_min) and featurizing sgRNAs (n_jobs_max). We recommend keeping n_jobs_min set to 1 or 2, as the APIs limit the amount of queries per hour.

design_df = pd.read_table('test_data/sgrna-designs.txt')
import multiprocessing
max_n_jobs = multiprocessing.cpu_count()

scored_designs = predict(design_df, tracr=['Hsu2013', 'Chen2013'], target=True,
                         n_jobs_min=2, n_jobs_max=max_n_jobs,
                         aa_seq_file='./test_data/target_data/aa_seqs.pq',
                         domain_file='./test_data/target_data/protein_domains.pq',
                         conservatin_file='./test_data/target_data/conservation.pq',
                         lite=False)
scored_designs.head()

Calculating sequence-based features


100%|██████████| 400/400 [00:05<00:00, 68.98it/s] 


Calculating sequence-based features


100%|██████████| 400/400 [00:01<00:00, 229.85it/s]
/opt/anaconda3/envs/rs3/lib/python3.8/site-packages/sklearn/base.py:310: UserWarning: Trying to unpickle estimator SimpleImputer from version 1.0.dev0 when using version 0.24.2. This might lead to breaking code or invalid results. Use at your own risk.
  warnings.warn(
/opt/anaconda3/envs/rs3/lib/python3.8/site-packages/sklearn/base.py:310: UserWarning: Trying to unpickle estimator Pipeline from version 1.0.dev0 when using version 0.24.2. This might lead to breaking code or invalid results. Use at your own risk.
  warnings.warn(

<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}

</style>

	Input	Quota	Target Taxon	Target Gene ID	Target Gene Symbol	Target Transcript	Target Reference Coords	Target Alias	CRISPR Mechanism	Target Domain	...	Picking Round	Picking Notes	RS3 Sequence Score (Hsu2013 tracr)	RS3 Sequence Score (Chen2013 tracr)	AA Index	Transcript Base	Missing conservation information	Target Score	RS3 Sequence (Hsu2013 tracr) + Target Score	RS3 Sequence (Chen2013 tracr) + Target Score
0	PSMB7	2	9606	ENSG00000136930	PSMB7	ENST00000259457.8	NaN	NaN	CRISPRko	CDS	...	0	Preselected	0.787640	0.559345	64	ENST00000259457	False	0.152037	0.939676	0.711381
1	PSMB7	2	9606	ENSG00000136930	PSMB7	ENST00000259457.8	NaN	NaN	CRISPRko	CDS	...	0	Preselected	-0.294126	-0.181437	46	ENST00000259457	False	0.064880	-0.229246	-0.116557
2	PRC1	2	9606	ENSG00000198901	PRC1	ENST00000394249.8	NaN	NaN	CRISPRko	CDS	...	0	Preselected	-0.043418	-0.220434	106	ENST00000394249	False	-0.063012	-0.106429	-0.283446
3	PRC1	2	9606	ENSG00000198901	PRC1	ENST00000394249.8	NaN	NaN	CRISPRko	CDS	...	0	Preselected	0.759256	0.453469	263	ENST00000394249	False	-0.126357	0.632899	0.327112
4	TOP1	2	9606	ENSG00000198900	TOP1	ENST00000361337.3	NaN	NaN	CRISPRko	CDS	...	1	NaN	0.424001	-0.197035	140	ENST00000361337	False	-0.234410	0.189591	-0.431445

5 rows × 68 columns

Here are the details for the keyword arguments of the above function

• tracr - tracr to calculate scores for. If a list is supplied instead of a string, scores will be calculated for both tracrs
• target - boolean indicating whether to calculate target scores
• n_jobs_min, n_jobs_max - number of cpus to use for parallel computation
• aa_seq_file, domain_file, conservatin_file - precalculated parquet files. Optional inputs as these features can also be calculated on the fly
• lite - boolean indicating whether to calculate lite target scores

By listing both tracrRNAs tracr=['Hsu2013', 'Chen2013'] and setting target=True, we calculate 5 unique scores: one sequence score for each tracr, the target score, and the sequence scores plus the target score.

We can compare these predictions against the observed activity from GeckoV2

gecko_activity = pd.read_csv('test_data/Aguirre2016_activity.csv')
gecko_activity.head()

<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}

</style>

	sgRNA Sequence	sgRNA Context Sequence	Target Gene Symbol	Target Cut %	avg_mean_centered_neg_lfc
0	AAAAAACTTACCCCTTTGAC	AAAAAAAAAACTTACCCCTTTGACTGGCCA	CPSF6	22.2	-1.139819
1	AAAAACATTATCATTGAGCC	TGGCAAAAACATTATCATTGAGCCTGGATT	SKA3	62.3	-0.793055
2	AAAAAGAGATTGTCAAATCA	TATGAAAAAGAGATTGTCAAATCAAGGTAG	AQR	3.8	0.946453
3	AAAAAGCATCTCTAGAAATA	TTCAAAAAAGCATCTCTAGAAATATGGTCC	ZNHIT6	61.7	-0.429590
4	AAAAAGCGAGATACCCGAAA	AAAAAAAAAGCGAGATACCCGAAAAGGCAG	ABCF1	9.4	0.734196

gecko_activity_scores = (gecko_activity.merge(scored_designs,
                                              how='inner',
                                              on=['sgRNA Sequence', 'sgRNA Context Sequence',
                                                  'Target Gene Symbol', 'Target Cut %']))
gecko_activity_scores.head()

<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}

</style>

	sgRNA Sequence	sgRNA Context Sequence	Target Gene Symbol	Target Cut %	avg_mean_centered_neg_lfc	Input	Quota	Target Taxon	Target Gene ID	Target Transcript	...	Picking Round	Picking Notes	RS3 Sequence Score (Hsu2013 tracr)	RS3 Sequence Score (Chen2013 tracr)	AA Index	Transcript Base	Missing conservation information	Target Score	RS3 Sequence (Hsu2013 tracr) + Target Score	RS3 Sequence (Chen2013 tracr) + Target Score
0	AAAACTGGCACGACCATCGC	CCGGAAAACTGGCACGACCATCGCTGGGGT	PSMB7	16.4	-1.052943	PSMB7	2	9606	ENSG00000136930	ENST00000259457.8	...	0	Preselected	-0.294126	-0.181437	46	ENST00000259457	False	0.064880	-0.229246	-0.116557
1	AAAAGATTTGCGCACCCAAG	TAGAAAAAGATTTGCGCACCCAAGTGGAAT	PRC1	17.0	0.028674	PRC1	2	9606	ENSG00000198901	ENST00000394249.8	...	0	Preselected	-0.043418	-0.220434	106	ENST00000394249	False	-0.063012	-0.106429	-0.283446
2	AAAAGTCCAAGCATAGCAAC	CGGGAAAAGTCCAAGCATAGCAACAGGTAA	TOP1	6.5	0.195309	TOP1	2	9606	ENSG00000198900	ENST00000361337.3	...	0	Preselected	-0.294127	-0.022951	50	ENST00000361337	False	-0.354708	-0.648835	-0.377659
3	AAAGAAGCCTCAACTTCGTC	AGCGAAAGAAGCCTCAACTTCGTCTGGAGA	CENPW	37.5	-1.338209	CENPW	2	9606	ENSG00000203760	ENST00000368328.5	...	0	Preselected	-0.667399	-0.308794	34	ENST00000368328	False	0.129285	-0.538114	-0.179509
4	AAAGTGTGCTTTGTTGGAGA	TACTAAAGTGTGCTTTGTTGGAGATGGCTT	NSA2	60.0	-0.175219	NSA2	2	9606	ENSG00000164346	ENST00000610426.5	...	0	Preselected	-0.402220	-0.622492	157	ENST00000610426	False	-0.113577	-0.515797	-0.736069

5 rows × 69 columns

Since GeckoV2 was screened with the tracrRNA from Hsu et al. 2013, we'll use these scores sequence scores a part of our final prediction.

plt.subplots(figsize=(4,4))
gpplot.point_densityplot(gecko_activity_scores, y='avg_mean_centered_neg_lfc',
                         x='RS3 Sequence (Hsu2013 tracr) + Target Score')
gpplot.add_correlation(gecko_activity_scores, y='avg_mean_centered_neg_lfc',
                       x='RS3 Sequence (Hsu2013 tracr) + Target Score')
sns.despine()

Predictions on the fly

You can also make predictions without pre-querying the target data. Here we use example designs for BCL2L1, MCL1 and EEF2.

design_df = pd.read_table('test_data/sgrna-designs_BCL2L1_MCL1_EEF2.txt')

scored_designs = predict(design_df,
                         tracr=['Hsu2013', 'Chen2013'], target=True,
                         n_jobs_min=2, n_jobs_max=8,
                         lite=False)
scored_designs

Calculating sequence-based features


100%|██████████| 849/849 [00:06<00:00, 137.86it/s]


Calculating sequence-based features


100%|██████████| 849/849 [00:02<00:00, 321.44it/s]


Getting amino acid sequences


100%|██████████| 1/1 [00:00<00:00,  1.77it/s]


Getting protein domains


100%|██████████| 3/3 [00:00<00:00, 899.29it/s]


Getting conservation


100%|██████████| 3/3 [00:00<00:00, 10.67it/s]
/opt/anaconda3/envs/rs3/lib/python3.8/site-packages/sklearn/base.py:310: UserWarning: Trying to unpickle estimator SimpleImputer from version 1.0.dev0 when using version 0.24.2. This might lead to breaking code or invalid results. Use at your own risk.
  warnings.warn(
/opt/anaconda3/envs/rs3/lib/python3.8/site-packages/sklearn/base.py:310: UserWarning: Trying to unpickle estimator Pipeline from version 1.0.dev0 when using version 0.24.2. This might lead to breaking code or invalid results. Use at your own risk.
  warnings.warn(

<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}

</style>

	Input	Quota	Target Taxon	Target Gene ID	Target Gene Symbol	Target Transcript	Target Alias	CRISPR Mechanism	Target Domain	Reference Sequence	...	Picking Round	Picking Notes	RS3 Sequence Score (Hsu2013 tracr)	RS3 Sequence Score (Chen2013 tracr)	AA Index	Transcript Base	Missing conservation information	Target Score	RS3 Sequence (Hsu2013 tracr) + Target Score	RS3 Sequence (Chen2013 tracr) + Target Score
0	EEF2	5	9606	ENSG00000167658	EEF2	ENST00000309311.7	NaN	CRISPRko	CDS	NC_000019.10	...	NaN	Outside Target Window: 5-65%	0.907809	0.769956	666	ENST00000309311	False	-0.115549	0.792261	0.654408
1	EEF2	5	9606	ENSG00000167658	EEF2	ENST00000309311.7	NaN	CRISPRko	CDS	NC_000019.10	...	NaN	BsmBI:CGTCTC; Outside Target Window: 5-65%	0.171870	0.040419	581	ENST00000309311	False	-0.017643	0.154226	0.022776
2	EEF2	5	9606	ENSG00000167658	EEF2	ENST00000309311.7	NaN	CRISPRko	CDS	NC_000019.10	...	1.0	NaN	1.393513	0.577732	107	ENST00000309311	False	0.172910	1.566422	0.750642
3	EEF2	5	9606	ENSG00000167658	EEF2	ENST00000309311.7	NaN	CRISPRko	CDS	NC_000019.10	...	1.0	NaN	0.904446	0.008390	406	ENST00000309311	False	0.121034	1.025480	0.129424
4	EEF2	5	9606	ENSG00000167658	EEF2	ENST00000309311.7	NaN	CRISPRko	CDS	NC_000019.10	...	1.0	NaN	0.831087	0.361594	546	ENST00000309311	False	0.036041	0.867128	0.397635
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
844	MCL1	5	9606	ENSG00000143384	MCL1	ENST00000369026.3	NaN	CRISPRko	CDS	NC_000001.11	...	NaN	Off-target Match Bin I matches > 3; Spacing Vi...	-0.792918	-0.663881	52	ENST00000369026	False	-0.299583	-1.092501	-0.963464
845	MCL1	5	9606	ENSG00000143384	MCL1	ENST00000369026.3	NaN	CRISPRko	CDS	NC_000001.11	...	NaN	Outside Target Window: 5-65%; poly(T):TTTT	-1.920374	-1.819985	5	ENST00000369026	False	-0.003507	-1.923881	-1.823491
846	MCL1	5	9606	ENSG00000143384	MCL1	ENST00000369026.3	NaN	CRISPRko	CDS	NC_000001.11	...	NaN	Spacing Violation: Too close to earlier pick a...	-1.101303	-1.295640	24	ENST00000369026	False	-0.285485	-1.386788	-1.581125
847	MCL1	5	9606	ENSG00000143384	MCL1	ENST00000369026.3	NaN	CRISPRko	CDS	NC_000001.11	...	NaN	Spacing Violation: Too close to earlier pick a...	-0.617431	-0.621436	30	ENST00000369026	False	-0.312348	-0.929779	-0.933784
848	MCL1	5	9606	ENSG00000143384	MCL1	ENST00000369026.3	NaN	CRISPRko	CDS	NC_000001.11	...	NaN	On-Target Efficacy Score < 0.2; Spacing Violat...	-0.586811	-0.664130	30	ENST00000369026	False	-0.312348	-0.899159	-0.976478

849 rows × 61 columns

We see that the predict function is querying the target data in addition to making predictions.