DESKGEN AI is the result of four years of research to determine biological variables influencing CRISPR guide design.
Running our brand-new Dunne 2017 scoring algorithm, alongside proprietary implementations of -omics pipelines and other bioinformatic tools, DESKGEN AI considers thousands of variables and makes smart decisions to improve activity, reduce experimental bias and select the best guides for your CRISPR library.
Dunne 2017 simulates your experiment and selects the top-ranked guides for every target that you need to investigate.
Compared with other top-performing algorithms, our approach is 32% more accurate in assessing the relative predicted performance of any given guide for each target gene.
With superior prediction of absolute performance of any given guide versus other algorithms, DESKGEN AI can adjust the total number of guides necessary per target, to ensure phenotypic readout and keep costs down by ignoring redundant guides that ramp up overall library cost.
DESKGEN AI was built with over 4,000,000 data points gathered from large-scale CRISPR experiments.
This number grows as we continue to deliver industry-leading CRISPR libraries and iteratively improve our algorithms, filters and features with carefully-selected machine learning models for optimal guide RNA design.