Accepted_test

Methods and Algorithms: We developed a mathematical model to predict the temporal dynamics of crRNA levels during the activation of the CRISPR-Cas system upon infection of a bacterial cell by foreign DNA. This model, utilizing principles of statistical thermodynamics and non-linear dynamics, incorporates known regulatory features of the system as well as our proposed mechanism of combined activation, by: 1) derepression of the cas genes promoter due to sequestration of H-NS by foreign DNA, and 2) activation of cas genes expression through positive feedback performed by LeuO and BglJ-RcsB. We systematically searched for parameter combinations at which the increase in the crRNA amount, indicative of the activation of the CRISPR-Cas system, can occur during the first 30 min upon entry of the foreign, AT-rich DNA. To identify the parameters exerting the most significant influence on eliciting this response from the system, we applied the Random Forest machine learning technique. Additionally, we conducted a bioinformatics analysis where 16,388 viruses were matched to their respective host bacteria harboring CRISPR-Cas using the Virus-Host DB database and tools for identifying CRISPR arrays in bacterial genomes. For each bacterial genus, we assessed the disparity in genomic AT content between bacterial and viral DNA.