Welcome to the CRISPRAnalyzeR Github page!
CRISPRAnalyeR is a web-based, interactive suite for the analysis and documentation of pooled CRISPR Screens.
See CRISPRAnalyzeR in action!
Check out the Wiki Page or scroll down for more information and help
See our Manuscript on biorxiv.org http://biorxiv.org/content/early/2017/02/20/109967
You can use the provided CRISPRAnalyzeR web-service or download the suite for installation within your lab.
CRISPRAnalyzeR has been developed to provide a fully-interactive and exploratory analysis of pooled CRISPR Screens with user experience in mind. You can analyse your screen using 8 different analysis methods as well as perform gene annotation, gene set analysis and get detailed information about your sgRNAs - all in a convenient web-browser interface.
All you need is your sequencing data and the a file describing your pooled CRISPR screen library (we provide you with the most common ones) - and you can go from rawdata to potential followup candidates
CRISPRAnalyzeR uses a guided-analysis approach. This means you will be guided through the analysis.
CRISPAnalyzeR consists of four sections:
- Screen and Sequencing Quality Estimation
- Hit Calling using multiple published hit calling algorithms
- Hit Confirmation which includes information from up to 26 external data ressources
- Download of the interactive report that provides a comprehensive documentation of your screen and analysis
CRISPRAnalyzeR assists you with the analysis of your screening data. It contains 4 different sections, each offering interactive visualizations and tables.
CRISPRAnalyzeR combines the power of several CRISPR Analysis Workflows and incorporates them into a streamlined and convenient workflow.
You run one analysis and get the information of up to 8 different analysis workflows
We implemented multiple available analysis workflows
- DESeq2 (based on gene-level read counts)
- MAGeCK
- sgRSEA
- edgeR
- BAGEL
- ScreenBEAM
- Mann-Whitney Test
Use up to 26 external data ressources to enrich information about your favourite candidate - all information is available within the application and can be saved to your report for documentation
Finally you can document the screening data and analysis using the interactive report.
You can find an example report HERE.
You can also check our live demo, which you can also use to analyse your screening data.
CRISPRAnalyzeR can be downloaded as as source code and is also available for a platform-independent installation as pre-configured docker image.
Please have a look at the installation tutorials below, which will assist you with the installation
We encourage users to obtain the pre-configured application which is described below.
The idea of installing CRISPRAnalyzeR is to provide a single installation within a Lab/Institute, so that everyone can access it via the web browser.
However, you can also also install CRISPRAnalyzeR on your local machine.
CRISPRAnalyzeR is based on R Shiny and uses many different R packages and tools which are listed separately. For a source code installation, we recommend the use of Ubuntu.
| Type | Source Code Installation | Docker Container Installation |
|---|---|---|
| Operating System | Ubuntu | Any supported OS by Docker |
| CPU | Dual Core (Quad Core recommended) | Dual Core (Quad Core recommended) |
| RAM | 8 GB | 8GB |
| HDD | 512 GB (SSD recommended) | 512 GB (SSD recommended) |
| Additional Software Packages | See list below! | All included in container |
CRISPRAnalyzeR is published under the GPL-2 license and is free for non-commercial use only.
While CRISPRAnalyzeR does not require an additional license for commercial use itself, some included tools strictly require additional licensing.
Please note that Highcharts, the COSMIC database and the Enrichr API access require additional licensing for commercial use.
The authors of this application are not responsible for licensing or license issues - please contact the corresponding companies directly..
Highcharts Licensing Options
https://shop.highsoft.com/
Please also have a look here: jbkunst/highcharter#254 (comment)
COSMIC Licensing Options https://cancer.sanger.ac.uk/cosmic/license
ENRICHR Licensing Options http://www.ip.mountsinai.org/
It is your responsibility to obtain all required licenses in case of commercial usage!
Version 1.141 (latest)
- adapted to changes in ENSEMBL database structure
Version 1.14
- stability enhancements in report generation
- fixed some UI typos
- prepared basis for new features
- added new parameter: max_upload -> sets the maximum allowed upload limit
- introduced Kitematic-based installation
Version 1.13
- fixed some typos
- fixed report generation issue with larger datasets
OLDER VERSIONS are not available anymore.
Installation Tutorial
This installation will use a tool called Kitematic to offer a user interface to start, stop or change settings of the CRISPRAnalyzeR.
In principal, you first install and start docker, then install and start kitematic and finally download CRISPRAnalyzeR.
Individual Steps
-
Download the Docker Installer for your operating system from the Docker Website
-
Install the downloaded file
-
Start Docker on your machine (e.g. by double clicking on the Docker icon on windows or Mac). A small docker symbol in the taskbar will tell you that docker is ready.
-
Download the Docker Toolbox from the docker website
-
Install the Docker Toolbox
-
Start Kitematic, which comes with the docker toolbox
-
Search for CRISPRAnalyzeR
ATTENTION In case you will see a red error message like
Could not connect to xxxx, this is an indication that you are behind a proxy server which blocks the access!
If this is the case, please use a network connection without a proxy OR follow the Installation Procedure for Advanced Users -
Select CREATE to install the latest stable version of CRISPRAnalyzeR to install CRISPRAnalyzeR
-
OPTIONAL If required, you can adjust a couple of parameters, such as a proxy server.
All parameters are described in more detail below
To adjust these parameters, click on the Settings tab on the upper right corner:
-
Hit the START button to start CRISPRAnalyzeR
-
Copy the URL to the web browser to access CRISPRAnalyzeR as shown in this picture:
-
Now you successfully installed and started CRISPRAnalyzeR!
Hit the start, stop or restart button.
As mentioned below, CRISPRAnalyzeR can be adjusted with several parameters.
With the user-interface provided by Kitematic, you can easily adjust all parameters from the Settings section.
Open Kitematic and click on the CRISPRAnalyzeR image - you will see a SETTING button on the top right.
-
Search for CRISPRAnalyzeR in the Kitematic UI repository
-
Click on the three dots for more options
-
Click on "Selected version"
-
Select the version you want
-
Click on the X at the bottom right
-
Click CREATE
-
Download the Docker Installer for your operating system from the Docker Website
-
Install the downloaded file
-
Start Docker on your machine (e.g. by double clicking on the Docker icon on windows or Mac). A small docker symbol in the taskbar will tell you that docker is ready.
-
Open a terminal or command line (macOS: Terminal; Windows: cmd.exe)
-
Download and run the CRISPRAnalyzeR directly from the online repository (without additional settings)
docker run --rm -p 80:3838 boutroslab/crispranalyzer:latestIt is important to keep the 80:3838 as this tells the software how to access it via the browser.
By default, CRISPRAnalyzeR can be accessed by the web-browser using http://localhost/CRISPRAnalyzeR.If you want to run a specific version, just replace the
latestwith the specific version numberdocker run --rm -p 80:3838 boutroslab/crispranalyzer:1.08All pre-configured versions are listed at the Docker Hub.
-
Familiarize with the parameters you can use to start the CRISPRanalyzeR - they offer proxy settings or additional databases and local sgRNA re-evaluation.
-
Access CRISPRAnalyzeR using your web-browser - http://localhost/CRISPRAnalyzeR
CRISPRAnalyzeR re-evaluates every sgRNA during the analysis process. Thus, it needs to map each sgRNA against the reference genome.
This can be performed locally (does not require a fast internet connection) or via e-crisp.org (requires fast internet connection >10 mbit).
By default, CRISPRAnalyzeR uses e-crisp.org to re-evaluate your sgRNAs. If you want to perform local sgRNA re-evaluation, please see below.
In case you would like to update CRISPRAnalyzer to the latest version, just type
docker pull boutroslab/crispranalyzer:latest
to download the latest version and then run it as described above:
docker run --rm -p 80:3838 boutroslab/crispranalyzer:latest
You can start and stop the CRISPRanalyzeR using docker.
CRISPRAnalyzeR has been designed to be installed once on a machine and then accessed via a webbrowser. Therefore, you can adjust all parameters during the start of the CRISPRAnalyzeR.
| Parameter | Meaning | Default Value | Accepted Values |
|---|---|---|---|
| websockets_behind_proxy | use Websocket protocol | TRUE | TRUE or FALSE |
| verbose_logfiles | output of log files | TRUE | TRUE or FALSE |
| COSMIC_database | Filename of the COSMIC Mutant database (CosmicMutantExport.tsv) | NULL | CosmicMutantExport.tsv |
| EnrichR | Whether to enable or disable the Enrichr API access | TRUE | TRUE or FALSE |
| EnrichR_URL | URL to the Enrichr API | http://amp.pharm.mssm.edu/Enrichr/ | Any URL |
| max_upload | Maximum size of uploaded files in Megabytes | 4096 | any number |
| bowtie_threads | Number of bowtie2 threads for mapping | 2 | any number, must be equal or smaller the number of CPU cores |
| proxy_url | URL to your Proxy server | NULL | URL or NULL to inactivate |
| proxy_port | Proxy server Port | NULL | Port number of NULL to inactivate |
In Kitematic, just go to the settings tab and adjust the options.
Be sure to hit the save button and start/restart CRISPRAnalyzeR.
All parameters can be set during the start by the following:
docker run --rm -e PARAMETER1 -e PARAMETER2 -e PARAMETER3 -p 80:3838 boutroslab/crispranalyzer:latestthis means you always need to add -e in front of the parameters, e.g.:
-e websockets_behind_proxy=TRUE -e verbose_logfiles=TRUE -e bowtie_threads=4e.g.
docker run --rm -e bowtie_threads=4 -e proxy_url="http://thisismyproxy.com" -e proxy_port=80 -p 80:3838 boutroslab/crispranalyzer:latestFor a source code installation you will need to have docker installed, as this will build your own docker image out of the source code and additional dependencies.
| Software | Version | Link |
|---|---|---|
| Bowtie2 | 2.29 | http://bowtie-bio.sourceforge.net/bowtie2/index.shtml |
| CRISPR ReEvaluation Tool | Latest | https://github.com/boutroslab/Supplemental-Material |
| PERL | 5 | https://www.perl.org/ |
| Python | 2.7.11 | https://www.python.org/ |
| Python Scipy | latest | https://www.scipy.org/ |
Navigate to the folder where this dockerfile is located. This will be the folder that is used to create the docker image file.
In general, CRISPRAnalyzeR source code will be retrieved by cloning. In principle, this can be changed to use any fork of CRISPRAnalzyeR, as long as the file structure remains the same!
Start docker and make sure you have set a proxy, if required.
First download the following dependencies to the directory containing this dockerfile
wget https://s3.amazonaws.com/rstudio-shiny-server-os-build/ubuntu-12.04/x86_64/shiny-server-1.5.2.837-amd64.deb
wget http://downloads.sourceforge.net/project/bowtie-bio/bowtie2/2.2.9/bowtie2-2.2.9-linux-x86_64.zip
wget https://sourceforge.net/projects/bowtie-bio/files/bowtie/1.2.0/bowtie-1.2-linux-x86_64.zip
wget http://downloads.sourceforge.net/project/mageck/0.5/mageck-0.5.5.tar.gzNow clone the CRISPRAnalyzeR project into the caR-release-git folder and sync the source to caR-release. The content of caR-release is the source code of CRISPRAnalyzeR and will be used for building the docker image.
mkdir ./caR-release
git clone git@github.com:boutroslab/CRISPRAnalyzeR.git ./caR-release-git
rsync -rav caR-release-git/ caR-release/source --exclude=.gitFinally build the image, which takes 1-3 hours depending on the computer and network speed.
docker build -t CRISPRAnalyzeR .docker run --rm -p 80:3838 CRISPRAnalyzeRCheck it by opening a browser tab and navigating to
http://localhost/CRISPRAnalyzeR
If this works you have successfully created a CRISPRAnalyzeR docker image!
Please note: COSMIC Database access only works in the commmand line installation.
The COSMIC database can be found here. By default, we do not provide the COSMIC database due to licensing compatibility. In case you want to use the COSMIC database, please proceed as follows.
- Visit the COSMIC Database Website
- If you aim for a commercial use, please see the COSMIC Licensing Information Page
- Head to the COSMIC download section
- Download the COSMIC Mutation Data database file
- Extract the CosmicMutantExport.tsv.gz file to the a database folder DATABASEFOLDER that you can define on your own
- Tell CRISPRAnalyzeR during the start that you have an external folder with your database files using the -v parameter -v DATABASEFOLDER:/srv/shiny-server/CRISPRAnalyzeR/database and that you would like to have the COSMIC database included via the parameter -e COSMIC_database="CosmicMutantExport.tsv"
Please replace DATABASEFOLDER byt the full path to the directory where you have placed the CosmicMutantExport.tsv!
docker run --rm -v *DATABASEFOLDER*:/srv/shiny-server/CRISPRAnalyzeR/database -e COSMIC_database="CosmicMutantExport.tsv" -p 80:3838 boutroslab/crispranalyzer:latestPlease note that the COSMIC database is loaded during the analysis procedure and requires 1 GB of RAM.
Enrichr offers API access for a gene set analysis. By default, CRISPRAnalyzeR has the Enrichr API access ENABLED. You can DISBALE the Enrichr API access during the start by setting the paratemer -e disable_EnrichR=TRUE
docker run --rm -e disable_EnrichR=TRUE -p 80:3838 boutroslab/crispranalyzer:latestPlease not that you require a license for commercial use. A license can be obtained by contacting the Mount Sinai Technology Development.
More information can be found at the Enrichr Help Page.
Please note: local sgRNA re-evaluation is only supported in the command line installation, not in the Kitematic installation
CRISPRAnalyzeR re-evaluated your sgRNAs during the analysis process. By default, this is done by sending the sgRNAs to www.e-crisp.org and downloading the files again. However, you can also perform a local re-evaluation of your sgRNAs!
You can download the human reference genome (or mus musculus) and tell CRISPRAnalyzeR to perform the re-evaulation locally on your computer. For this, you require at least 60GB of disk space and a pretty fast computer.
Step 1: Download the reference genome
Download the reference genome you need
Step 2: Extract the files to a folder (DATABASEFOLDER) of your desire
Extract the downloaded file using gunzip (macOS/Linux) or Zip (Windows) to a DATABASEFOLDER of your desire.
Please note that you need to know the exact path to the DATABASEFOLDER!
e.g. /home/user1/databases
If you have already setup the COSMIC Database, please make sure you extract the file into that same folder.
Please note: The folder structure MUST be kept, when you extract the file into
e.g. /home/user1/databases
this is your DATABASEFOLDER
it will create a folder structure like that:
e.g. /home/user1/databases/data/scripts/crispr_databases/homo_sapiens
Step 3: start CRISPRAnalyzeR and tell it where you files are Using the command line, start CRISPRAnalyzeR and provide the database path DATABASEPATH - so that CRISPRAnalyzeR know where to look for your data.
Please note: if you have setup the COSMIC database already, just extract the files in the same folder and you do not need to do anything in addition
Again, replace DATABASEFOLDER with your absolute path.
docker run --rm -v *DATABASEFOLDER*:/srv/shiny-server/CRISPRAnalyzeR/database -p 80:3838 boutroslab/crispranalyzer:latestYOUTUBE VID HERE
CRISPRAnalyzeR comes with read count sample data that can be accessed from the help on the data upload section.
Moreover, we offer additional sample data that you can download from our website:
- Drug Resistance Screen Data from CRISPR Library Designer
- Essential Genes Screen Data using the TKOv1 library
This data was published before in F. Heigwer*, T. Zhan*, M. Breinig, J. Winter, D. Brügemann, S. Leible, M. Boutros, CRISPR library designer (CLD): software for multispecies design of single guide RNA libraries, Genome Biol., 2016, DOI:10.1186/s13059-016-0915-2
You can download pre-made read count files for data analysis from here.
These include four read count text files as well as the sgRNA library file in FASTA format.
You can directly use these files with the default parameters within CRISPRAnalyzeR.
You can download FASTQ Sequencing files from here.
These include four NGS raw data files in compressed fastq format as well as the sgRNA library file in FASTA format.
You can directly use these files with the default parameters within CRISPRAnalyzeR.
This data was published in Steinhart,Z. et al. (2016) Genome-wide CRISPR screens reveal a Wnt-FZD5 signaling circuit as a druggable vulnerability of RNF43-mutant pancreatic tumors. Nat. Med.
You can download raw read count data including a TKO sgNRA library FASTA file from here
In order to use these files, you need to adjust the gene identifier from Ensembl Gene ID to HGNC Symbol in the Analysis Settings.
CRISPRAnalyzeR offers pre-made sgRNA library files in FASTA format for use. You can use them along with your read count (please see the format) or raw NGS sequencing files (.fastq.fz).
| Library Name | Lab | Pubmed ID | Addgene | Download |
|---|---|---|---|---|
| CLD Benchmarking | Boutros | 27013184 | NA | FASTA |
| Gecko V2 | Zhang | 25075903 | here | A+B FASTA A FASTA B FASTA |
| Gecko V2 MOUSE | Zhang | 25075903 | here | A FASTA B FASTA |
| Torronto KnockOut Library (TKOv1) | Moffat | 26627737 | here | 90K FASTA & 85K FASTA |
| Brunello | Doench | 26780180 | here | FASTA |
| CRISPRa / CRISPRi | Weissmann | 25307932 | here | not available yet |
| Human Lentiviral sgRNA library high cleavage activity | Sabatini | 26472758 | here | 185K FASTA |
| Human Lentiviral sgRNA sub libraries | Sabatini | 24336569 | here | not available yet |
If you use FASTQ sequencing files directly, CRISPRAnalyzeR requires the use of a so-called 'regular expression' to extract the sgRNA target sequence from your sequencing data.
However, the CRISPRAnalzyeR comes with some pre-defined settings for commonly used screening libraries and their vector systems. In case you don't know which is the best setting for your screen, please do not hesitate and create an issue or write me an email.
| Plasmid Name | Lab | Addgene ID | Regular Expression |
|---|---|---|---|
| Lenticrisp V2 | Feng Zhang | 52961 | Default ACC(.{20,21})G |
| Lentiguide (Puro) | Feng Zhang | 52963 | Default ACC(.{20,21})G |
| Human Lentivirus Library V1 | Haoquan Wu | 69763 | GTTT(.{20})G |
| pLCKO (Moffat TKO) | Moffat | 73311 | ACCG(.{20,21})G |
| pU6-sgRNA EF1Alpha-puro-T2A-BFP (CRISPRa/i) | Weissman | 60955, 62217, 60956 | GTTG(.{20})G |
CRISPRAnalyzeR requires you to upload a sgRNA library FASTA file which contains all sgRNAs of your screen. Moreover, you can either use your sequencing data directly as gzipped FASTQ files (.fastq.gz) or already calculated, non-normalized, read count files (.txt).
A sgRNA library file must be in FASTA format and include the unique sgRNA identifier as well as the target sequence in 5'->3' direction. Please make sure that only the exact target sequence is used here.
>ENSG00000006042_0_4299.561
GGAGCCCTCTGAGTTAGAAC
>ENSG00000006042_5_4299.588
GAAGATGCCTCGTAAGGCCA
>ENSG00000006042_6_4299.588
GAGATGCCTCGTAAGGCCAT
Moreover, the gene identifier needs to be a part of the unique sgRNA identifier as shown here
so that CRISPRAnalyzeR knows which sgRNA belongs to which gene.
Please note that all sgRNA identifiers need to be unique.
Next Generation Sequencing files usually are provided in FASTQ format:
@M01100:47:000000000-AH40C:1:1101:12289:2057 1:N:0:4
AACACCGTCAGTGTGCTTGCCCCACTGTTTTAGAGCTAGAAATAGCAAGTT
+
GGGGGGGGGGDGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGG
@M01100:47:000000000-AH40C:1:1101:8758:2058 1:N:0:4
AAACACCGGTTTTGAAACTGCGGACACGTTTAGAGCTAGAAATAGCAAGTTA
+
GGGGGGGEGGGGGGGGGDGGFGGGEEGGFFGGFFCFFF=AFF<CEFFF@EFE
For easier use and faster data handling, CRISPRAnalyzeR required gzipped FASTQ files (.fastq.gz).
FASTQ files are usually provided in a gzipped format by the sequencing facility.
CRISPRAnalyzeR requires a separate read count file for each sample, which needs to be tab-separated.
sgRNA Count
ENSG00000053900_GAAAGCAATGAGATCCCGCT 28
ENSG00000053900_GAAGCAATGAGATCCCGCTT 62
ENSG00000053900_GAAGCGGGATCTCATTGCTT 92
The first column describes which sgRNA identifier (must be the same as in the sgRNA library FASTA file) was used, the second column describes how many reads were present for this unique sgRNA.
Please note that all sgRNA identifiers need to be unique.
All available tutorials can be found on Youtube and are also available from within CRISPRAnalyzeR.
CRISPRAnalyzeR can provide you the log files generated, which can help you to identify potential issues and me to fix bugs or add new features.
In order to see the logs conveniently without struggeling into the running docker application itself, you can ask CRISPRAnalyzeR to export the logs in real-time to any folder using the start parameter -v in a similar way as it is done to include the COSMIC database or local sgRNA re-evaluation.
LOGFOLDER is a local folder you create to which CRISPRAnalyzeR will save and update the logs in real time.
Start CRISPRAnalyzeR using the -v command, e.g.
docker run --rm -v *LOGFOLDER*:/srv/shiny-server/CRISPRAnalyzeR/log -p 80:3838 boutroslab/crispranalyzer:latestFor some genome-wide screens, CRISPRAnalyzeR can abort the report generation due to limited amount of storage allocated for this operation. You can force docker to give more space to CRISPRAnalyzeR by increasing the ulimit setting (in kilobytes)
For 8 GB of the so called c stack space, you need to set the value of 8GB in Kb - 8.589.934.592 with --ulimit stack=8589934592:8589934592.
docker run --rm --ulimit stack=8589934592:8589934592 -v *LOGFOLDER*:/srv/shiny-server/CRISPRAnalyzeR/log -p 80:3838 boutroslab/crispranalyzer:latestA status page for the live demo availability can be found here.
Jan Winter - jan.winter@dkfz-heidelberg.de
Twitter: @winterj86