TangSoftwareLab
commented
1 year ago Hi Aleksandr, Mathematically CSS differs from DSS in two aspects: 1) CSS uses a log-logistic function while DSS uses a logistic function; 2) CSS calculates the AUC of the log10 scale of the dose-response curve while DSS uses the raw dose scales. Both CSS and DSS are variants of AUC-based methods, and our paper has cited them properly (citations 21-22). Best, Jing
IanevskiAleksandr
Dear Developers,
I have recently been reviewing the implementation of CSS metric in your SynergyFinder Plus package. I noticed that it bears a striking resemblance to the Drug Sensitivity Score (DSS) published in 2014 (https://www.nature.com/articles/srep05193). Specifically, CSS is defined as almost just an average of two DSS values (even with same normalizations applied), calculated with responses when another drug is fixed at its IC50.
Could you please clarify the novelty of the CSS metric compared to DSS? In particular, I am curious as to why the metrics are referred to as CSS1 and CSS2 in the manuscript, rather than DSS1 and DSS2 for drugs 1 and 2, respectively, given that calculation of DSS in the presence of another drug has been a common practice in many previous studies? Additionally, I would appreciate an explanation as to why the manuscript does not explicitly mention that CSS is a direct adaptation of DSS with only minor modifications. Providing this information could have been beneficial for reviewers in determining if the CSS metric is sufficiently novel to warrant distinct recognition.
Thank you for your attention to this matter, and I look forward to your response.
Sincerely,