Call HS regions from CRISPR tiling screen data and predict HS region from common protein features
Project description
Copyright (C) 2019, MD Anderson Cancer Center (whe3@mdanderson.org)
ProTiler
ProTiler is a novel computational method for fine-mapping of protein regions that are hyper-sensitive to CRISPR/Cas9 mediated gene knockouts from high-throughput tiling-sgRNA functional screens.
Also, ProTiler is able to predict HS regions for protein encoded by any given gene from other common protein features including conservation, domain annotation, secondary structures and PTMs distribution.
If you use ProTiler please cite the following paper we published on Nature Communications:
He et al. De novo identification of essential protein domains from CRISPR-Cas9 tiling-sgRNA knockout screens. Nat Commun 10, 4547(2019).
Installation
ProTiler is written in Python and R, Python>=2.7 and R>=3.5.0 is needed
Dependencies
Python Packages:
- matplotlib
- pandas
- numpy
- seaborn
R packages:
- breakfast
- stringr
Step1: Install Anaconda (highly recomended)
wget https://repo.continuum.io/archive/Anaconda2-2018.12-Linux-x86_64.sh
bash Anaconda2-2018.12-Linux-x86_64.sh
Step2: Install required packages
Install Python Packages with pip:
pip install matplotlib==2.2.3 pandas sklearn numpy seaborn
Install R packages in R IDE:
install.packages('breakfast')
install.packages('stringr')
Step3: Install ProTiler
Through git clone
git clone https://github.com/MDhewei/ProTiler-1.0.0.git
cd ProTiler-1.0.0
python setup.py install
Through pip
pip install protiler
Usage
ProTiler has two major functions:
1. Call: Call and visualize HS regions from CRISPR tiling screen data.
Protiler call take table file(.cvs or .txt) recording CRISPR tiling screen data as inpu.An example is shown as below:
Three colums are required:
-
Symbol: This column record the symbol of target gene, for example: 'CREBBP','ACTL6A'
-
AA: the amino acid position which certain sgRNAs cutting (what does this means?)
-
CRISPR score: the signals for each sgRNA, in the example file, z-scores in three different cell lines are used. User should select at least one column.
Arguments of the program:
Required arguments:
-
-i/--inputfile:
the file path to the input table recording tiling CRISPR sgRNA annotations and signals. .csv,.txt,.xlsx format are supported
-
-g/--gene_id:
the official symbol of target gene, for example: 'CREBBP','ACTL6A'
-
-s/--score_columns:
the column number(s) of input table that recording CRISPR knowckout scores
Optional arguments:
-
-o/--outputdir:
the directory name created in the current working directory to save output files, default='ProTilerOutput'
-
-f/--half_size:
The number of neiboring signals from each side selected to filter inefficient sgRNAs',default='5'
-
-t1/--threshold:
Threshold to supress the outliers among the signals',default='2'
-
-t2/--threshold2:
Threshold to detect changing points using TGUH method',default='1.5'
Example to run protiler call
protiler call -i sample.txt -g CREBBP -s 9,10,11 -o ProtilerOutput
Output
1. SegmentFile: A table record all the HS regions called by ProTiler for certain gene.
- AA.start: the start residue position of the segments called with TGUH
- AA.end: the end residue position of the segments called with TGUH
- n: the number of sgRNAs targeting the region
- m: the mean score of sgRNAs targeting the region
- is.HS.site: to judge whether the segment is a hyper-sensitive region
- length: the length of the segment
- Gene: the symbol of the target gene
2. Figure4Visualization: Figure presenting signals, HS regions and other protein annotations. For example:
2. Predict: Predict and visualize HS regions from common protein features.
Arguments of the program:
Required arguments:
-
-l/--gene_list:
A list of candidate genes for which you want to predict HS regions. eg: CREBBP,FAM122A,AURKB
Optional arguments:
-
-b1/--bandwidth1:
Bandwidth for PTMs kernel density estimation training
-
-b2/--bandwidth2::
Bandwideth for SIFT score kernel density estimation training
-
-o/--outputdir:
the directory name created to save output files
-
-m/--gamma:
The gamma parameter for SVM model,default='10'
-
-c/--penalty:
The penalty parameter for SVM model,default='0.01'
Example to run protiler predict
protiler predict -l CREBBP,FAM122A,SAMRCB1,AURKB -o ProtilerOutput
Output
1. PredictionTable: A table record all the features of target protein and SVM score/class at each residue postion.
2. Figure4Visualization: Figure presenting predicted HS regions and other protein annotations. For example:
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.