Help

What is CRISPR Designer?

CRISPR Designer is a web service dedicated to assist the design of guide RNA in CRISPR/Cas9 system. CRISPR Designer identify potential on-target sites and off-target sites by aligning the input sequences to the whole genome using an efficient binary alignment scheme. The efficacy of all identified target sites are then quantitatively assessing by Random Forest regression model. To maximize the on-target efficacy and minimize the off-target effect, a composite score are calculated and used to select optimized sgRNA designs. All the identified targets can be visualized in genome browser.

How do I use it?

1. Firstly, to use CRISPR designer you will need to upload a nucleotide sequence file in FASTA format.

An example for FASTA format:

>WASH7P
GAGATATTTTTAGAGACTGGACCTGAGGCCTCTGGAGGCTAC
>RP5-857K21.4
ATTCGATGTTGAAGCCTGAGACTAGTTCGGACTCCCTTTGAC

Note: Avoid using space or semicolon in your FASTA headers. If space or semicolon are used, CRISPR Designer will use the text before the first space or semicolon as header. For example, >example;1 will become >example. Also, please notice that the FASTA headers must be unique. Duplicate headers will be removed by the program.

2. Next, select the corresponding species. Currently, human and mouse are supported in our web service.

3. Select the number of optimized sgRNA that you want to export and click on the submit button to start the calculation. The current status of your job can be viewed in the console panel. When the calculation is done, you can click on the view button to access the results.

What does the sgRNA score mean?

The sgRNA score is a composite score that quantitatively represent how well a sgRNA design is likely to be highly active and specific. The score are calculated as the sum of on-target efficacy minus the sum of off-target efficacy. A higher score usually indicates a better sgRNA design.

How does the CRISPR Designer calculate on-target and off-target efficacy?

We calculate the target efficacy using three Random Forest based regression models. From the recently published literatures, we collected 1841 on-target sites, 847 single mismatch off-target sites and 184 multiple mismatch off-target sites. The efficacy of all those collected target sites are measured by in vivo experiments. Based on these in vivo efficacy, we trained an on-target regression model, a multiple mismatch off-target model and a single mismatch off-target model. When inputting a potential target sites, the sequence features are extracted and the efficacy are predicted by the corresponding regression model.

I'm having trouble with CRISPR Designer that isn't addressed on this page. What should I do?

If you are having trouble with CRISPR Designer please contact the two major authors: Dr. Jian Ren and Dr. Penghui Zhou. We will try resolve it.