Jupyter notebook powers reproducible and interoperable genomic research

Did you know that when targeting the same gene in different tissues/organs, you may have to use a different target site for the genome editing machinery (CRISPR-Cas9) to work optimally? Similar to Google maps suggesting an alternative route on your drive home, to avoid peak hour traffic.

To this end, the CSIRO team has developed GT-Scan2, which functions as a molecular “navigation system” that evaluates different genome-editing options and suggests the most effective solution for the experimental setup, organs, conditions, ect. It does this by taking advantage of the chromatin architecture, which can change between different experimental setups.

Hence running GT-Scan2 on two different cell-types and comparing the target-site predictions allows to identify sites that are universally active (active in both) or differentially active (active in one but not the other).

Using Jupyter and API calls to the GT-Scan2 web-service you can get accurate CRISPR-target site predictions directly into your notebook. This allows downstream analysis of the results, like finding target-sites that are active in one condition but inactive in another.

More in the Medium article.