Interpretation of synergy scores

[Lux91]

Hi, I'm new to the world of drug synergy experiments and I'm using synergyfinder on R to compute 3-drug synergy scores for a research project. I'm having some trouble trying to determine if a combination is considered synergistic or not under a specific model. For example, I know that the Highest Single Agent score points to synergy when above zero and to antagonism when below zero but I can't understand if other scores (Bliss, ZIP, Loewe) have the same direction or not in your implementation. Could you help me figure out for each score what is the threshold value (like zero is for HSA) and their directions (for example above threshold value = synergy and below = antagonism or vice versa)? Thank you in advance for your help :)

[jtang920]

Hi Lucas, Yes, all the four scores have the same reference values, i.e. 0. Note that they may not agree on the same directions. For example, it may be so that HSA score is above 0 while the other scores below 0. Ideally, all the scores should be positive to call it a concensus synergy. Best, Jing

From: Lucas @.> Sent: Tuesday, October 10, 2023 12:21 PM To: TangSoftwareLab/SynergyFinderR @.> Cc: Subscribed @.***> Subject: [TangSoftwareLab/SynergyFinderR] Interpretation of synergy scores (Issue #10)

Hi, I'm new to the world of drug synergy experiments and I'm using synergyfinder on R to compute 3-drug synergy scores for a research project. I'm having some trouble trying to determine if a combination is considered synergistic or not under a specific model. For example, I know that the Highest Single Agent score points to synergy when above zero and to antagonism when below zero but I can't understand if other scores (Bliss, ZIP, Loewe) have the same direction or not in your implementation. Could you help me figure out for each score what is the threshold value (like zero is for HSA) and their directions (for example above threshold value = synergy and below = antagonism or vice versa)? Thank you in advance for your help :)

— Reply to this email directly, view it on GitHubhttps://github.com/TangSoftwareLab/SynergyFinderR/issues/10, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ADFWHIEIYNC2ZYKDN2NXSDTX6UHQ3AVCNFSM6AAAAAA52AWQIWVHI2DSMVQWIX3LMV43ASLTON2WKOZRHEZTIOBWHE2DMNA. You are receiving this because you are subscribed to this thread.Message ID: @.***>

[Lux91]

Thank you so much for your clarification. Since I'm already here, I am also trying to understand the p-values that are stored in the $drug_pairs slot of an object after scoring. From my understanding it is similar in principle to the result of a one sample t-test where you test a set of data (scores) against a mean of 0 (null-hypothesis of non-interaction). Is my understanding correct? Do I need these p-values to be statistically significant in order to consider valid the score from which they were computed? Thank you again,

[jtang920]

Hi Lucas, Yes the idea is to have a permutation test against 0 (null hypothesis of non-interaction). The p-values need to be significant to call it a synergy. Best, Jing

From: Lucas @.> Sent: Thursday, October 19, 2023 2:13 PM To: TangSoftwareLab/SynergyFinderR @.> Cc: Tang, Jing @.>; Comment @.> Subject: Re: [TangSoftwareLab/SynergyFinderR] Interpretation of synergy scores (Issue #10)

Thank you so much for your clarification. Since I'm already here, I am also trying to understand the p-values that are stored in the $drug_pairs slot of an object after scoring. From my understanding it is similar in principle to the result of a one sample t-test where you test a set of data (scores) against a mean of 0 (null-hypothesis of non-interaction). Is my understanding correct? Do I need these p-values to be statistically significant in order to consider valid the score from which they were computed? Thank you again,

— Reply to this email directly, view it on GitHubhttps://github.com/TangSoftwareLab/SynergyFinderR/issues/10#issuecomment-1770640257, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ADFWHIG6KNH5XEUEYBLHZ7TYAEDMBAVCNFSM6AAAAAA52AWQIWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTONZQGY2DAMRVG4. You are receiving this because you commented.Message ID: @.***>