Bioinformatics · Python · MIT license

PrimedSherlock: Cas12 crRNA Design Tool

PrimedSherlock is a free, open-source Python tool that designs highly specific CRISPR-Cas12 and Cas13 crRNAs from RPA primer sets. It was published in BMC Bioinformatics in 2022.

Get it on GitHub Read the paper

What PrimedSherlock does

CRISPR-Cas12 diagnostic assays pair an isothermal amplification step, recombinase polymerase amplification (RPA), with a Cas12 crRNA that recognizes the amplified target. Picking crRNAs by hand is slow, and a poorly chosen crRNA can react with related, non-target sequences.

PrimedSherlock automates that choice. It starts from RPA primer sets, proposes crRNAs inside each amplicon, and screens them against off-target (background) sequences so that only highly specific crRNA and primer combinations are kept. In the published study, crRNAs it designed against West Nile virus, Zika virus, and all four dengue virus serotypes detected their targets with minimal off-target background.

How it works: from RPA primer sets to specific crRNAs

Primer sets in. PrimedSherlock reads the Output_Sets.csv primer sets generated by PrimedRPA, the recommended source of RPA primers. Hand-made primer pairs are also supported.
On-target and background sequences. It uses your on-target genomes (input.fna) and the PrimedRPA background database (background_db.fasta).
Off-target screening. Candidate crRNAs are checked with Cas-OFFinder, on the CPU or a GPU.
Results out. Each run writes [RunID]_FinalOutput.csv with the crRNA and primer sets that passed.

Install and run

Set up a Python environment with Biopython, pandas, and configparser, then clone the repository into a new folder:

gh repo clone JamesGerardMann/PrimedSherlock

Copy three files from your PrimedRPA run into that folder:

background_db.fasta
input.fna
Output_Sets.csv

Choose the CPU or GPU that Cas-OFFinder should use. From a command prompt in the PrimedSherlock folder, run Cas-Offinder.exe and note the ID of your device under “Available Device List”. In the included .py file, replace each G0 with that device ID (for example C1 for CPU 1). Save the file, then run script.bat.

Key parameters

MismatchThreshold: how many mismatches a crRNA target may have against the included strains, usually 0–3. Most published examples used 3; the Zika example used 5. If no crRNA sets are produced, raise this value.
Use_PrimedRPA: read a PrimedRPA Output_Sets.csv. Set to No to supply your own primer pairs.
Threadcount: CPU threads for Cas-OFFinder; more threads run faster.
Amplicon_Min_Size / Amplicon_Max_Size: amplicon size limits. Smaller amplicons amplify faster with RPA polymerase.
Max_Background_Identity: PrimedRPA’s off-target amplification threshold; lower is stricter.
Off_Target_Mismatch_Score: the minimum number of mismatches to off-target sequences for a crRNA to count as safe.
Contingency_crRNA_Utilization: experimental; leave it off.

Cite PrimedSherlock

Mann, J. G., & Pitts, R. J. (2022). PrimedSherlock: a tool for rapid design of highly specific CRISPR-Cas12 crRNAs. BMC Bioinformatics, 23, 428. https://doi.org/10.1186/s12859-022-04968-5

DOI 10.1186/s12859-022-04968-5 Free full text (PMC9569017) PubMed 36241974

Third-party software and support

PrimedSherlock incorporates parts of PrimedRPA and includes a Cas-OFFinder executable. It is released under the MIT license. Report bugs through the GitHub repository.