Last updated: 2023-06-15

Checks: 0 1

Knit directory: combiCART/analysis/

This reproducible R Markdown analysis was created with workflowr (version 1.7.0). The Checks tab describes the reproducibility checks that were applied when the results were created. The Past versions tab lists the development history.


Tracking code development and connecting the code version to the results is critical for reproducibility. To start using Git, open the Terminal and type git init in your project directory.


This project is not being versioned with Git. To obtain the full reproducibility benefits of using workflowr, please see ?wflow_start.


Section 1: Data pre-processing and quality control

In this part, the raw screening data were cleaned-up and processed. Quality controls, such as estimating incubation effect and edge effect, were also performed.

Section 2: Exploratory analysis analysis

In this part, I performed some preliminary analysis mainly focused on single agent effect, such as comparing control, NT only, CAR-T only effect; plotting dose-response curves of drug + NT and drug + CAR-T effect and compare effects of T-cell donors. Those can be used for sainity check and quality control

Section 3: Identify synergistic combination effect using Bliss independence model

In this part, combination effect was evaluated using Bliss independence model (a breif review of different models: https://www.sciencedirect.com/science/article/pii/S259025712200030X?via%3Dihub). Basically, synergy here means only if the observed combination effect (drug+CAR-T) is stronger than expected additive effect of drug only and CAR-T only, i.e one plus one larger than two, they will be considered as true synergies. Hypothesis testing could also be performed to derive P-values by treatment donors as biological replicates.

Section 3.1: Same as Section 3 but without toxic agents

This same analysis as above, but single agents (either drug or CAR-T) that are already very toxic (less than 20% remaining viability in all three donors) were excluded from further analysis.

Section 4: Identify additive combination effect using Highest single agent model

The analyses are very similar to the ones above. The difference is that the combination effect was evaluated by using Highest single agent model. (a brief review of different models: https://www.sciencedirect.com/science/article/pii/S259025712200030X?via%3Dihub). Basically, if a observed combination effect is stronger than the highest effect of either of drug or CAR-T alone, it will be considered as candidate. The model is less stringent than Bliss independence model, where only true synergism will be considered. This model will identify any beneficial effect (synergy + additive) of combination compared to single agent.

Section 4.1: Same as Section 4 but without toxic agents

This same analysis as above, but single agents (either drug or CAR-T) that are already very toxic (less than 20% remaining viability in all three donors) were excluded from further analysis.