CRISPR-Net: A Recurrent Convolutional Network Quantifies CRISPR Off-target Activities with Indels and Mismatches

This repository includes a recurrent convolutional neural network named CRISPR-Net for predicting the off-targets activities with insertions, deletions and mismatches in CRISPR/Cas9 gene editing. There are two command-line tools that can be used to quantify the off-target activities induced by CRISPR guide RNA. One is CRISPR_Net.py which can predict the off-target activities with indels and mismatches, the other is CRISPR_Net_Aggregate.py which aggregates the gRNA-target scores from CRISPR-Net into a single consensus off-target score.

PREREQUISITE

CRISPR-Net was conducted by Python 3.6 and Keras 2.2.4 (using TensorFlow 1.12 backend)

Following Python packages should be installed:

• scipy

• numpy

• pandas

• scikit-learn

• TensorFlow

• Keras

Usage

CRISPR-Net can run with:

python CRISPR_net.py [-h] input_file

positional arguments: input_file

optional arguments: -h, --help show the help message and exit

Example input file:

GAGT_CCGAGCAGAAGAAGAATGG,GAGTACCAAGTAGAAGAAAAATTT
GTTGCCCCACAGGGCAGTAAAGG,GTGGACACCCCGGGCAGGAAAGG
GGGTGGGGGGAGTTTGCTCCCGG,GTGTGGGGTAAATTTGCTCCCAG
GGGTGGGGGGAGTTTGCTCCAGG,AGGTGGGGTGA_TTTGCTCCAGG
GATGGTAGATGGAGACTCAGNGG,GGTAGGAAATGGAGAGGCAGAGG
GAGTCCGAGCAGAAGAAGAAAGG,GAGTTAGAGCAGAAGAAGAAAGG

Save it as 'input.txt'.

Now you can run CRISPR-Net as following:

$> python ./CRISPR_net.py input.txt
...

Then output will be generated:

• The first column of the output file indicates the on-target sequence,
• The second column of the output file indicates the off-target sequence,
• The third column is the off-taget score predicted by CRISPR-Net

and saved to ./CRISPR_net_results.csv:

on_seq,off_seq,CRISPR_Net_score
GAGT_CCGAGCAGAAGAAGAATGG,GAGTACCAAGTAGAAGAAAAATTT,0.000000e+00
GTTGCCCCACAGGGCAGTAAAGG,GTGGACACCCCGGGCAGGAAAGG,1.490116e-07
GGGTGGGGGGAGTTTGCTCCCGG,GTGTGGGGTAAATTTGCTCCCAG,6.078471e-01
GGGTGGGGGGAGTTTGCTCCAGG,AGGTGGGGTGA_TTTGCTCCAGG,3.451956e-01
GATGGTAGATGGAGACTCAGNGG,GGTAGGAAATGGAGAGGCAGAGG,5.029583e-05
GAGTCCGAGCAGAAGAAGAAAGG,GAGTTAGAGCAGAAGAAGAAAGG,9.269089e-01

---

CRISPR-Net-aggregate can run with:

python CRISPR_net_aggregate.py gRNA_offTargets.csv

positional arguments: gRNA_offTargets.csv

optional arguments: -h, --help show the help message and exit

Example input file:

off_target,Gene_mark,on_target
GAACTAGCCTTGTATCCCAGGGA,RP4-669L17.10,GACCTTGCATTGTACCCGAGGGG
GAACTAGCCTTGTATCCCAGGGA,RP5-857K21.4,GACCTTGCATTGTACCCGAGGGG
GTGTTTGCAATGTACCCGTGTTG,TTLL10-AS1,6,1173076,GACCTTGCATTGTACCCGAGTGG
GACCTGTGGTTGTTCCTGAGAGG,,GACCTTGCATTGTACCCGAGAGG
GCCCTTGGATTGGCCGCGAGGGC,,GACCTTGCATTGTACCCGAGGGG

Save it as 'gRNA_offTargets.csv'.

Now you can run CRISPR-Net-Aggregate as following:

$> python ./CRISPR_Net_aggregate.py gRNA_offTargets.csv
...

Then off-target aggregate score will be generated.

You can try CRISPR-Net-Aggregate on the our example file (input_example/aggregate_example_GACCTTGCATTGTACCCGAG.csv) by:

python CRISPR_net_aggregate.py ./input_example/aggregate_example_GACCTTGCATTGTACCCGAG.csv

CONTAINS:

• code/CRISPR_Net.py : Python script to run CRISPR-Net for predicting off-target activities with indels and mismathces

• code/CRISPR_Net_Aggregate.py : Python script to run CRISPR-Net-Aggregate that aggregates the gRNA-target scores from CRISPR-Net into a single consensus off-target score.

• code/evaluate_CRISPR_Net.py : Python script to retrain CRISPR-Net on any dataset in /data and evaluate its performance.