This application searches for CRISPR off-targets on an NVIDIA GPU. crispacuda will use a dedicated GPU to to load known CRISPR sequences into memory and then perform searches on these sequences.
In order for crispacuda to run it needs an index file containing an array of all known CRISPRs in a genome assembly. The files are in binary format and are created using the CRISPR-Analyser utility. Follow the Create Index instructions on how to create a binary index file to use in crispacuda.
crispacuda is run from the command line with the following options:
[-i <FILE>]- specify the index file to search (required)[-s <INT>]- the index to start the searches from[-n <INT>]- how many off-targets to calculate from the start[-m <INT>]- the maximum number of mismatches to records results for. Defaults to 4[-q]- do not report a list of off-targets[-d <INT>]- the GPU device to use[-z]- specify CRISPRs as strings rather than IDs[-h]- print a help message and GPU information