To do

[jhcreed]

Content team

Page one - tumor level:

• [ ] spatial plots color coding points by positivity
• [ ] summary plots of distributions of markers
• [x] ripleys K plots (along various values of R)

Page two - study level:

• [x] violin plots of overall distributions
• [x] heatmaps with clinical annotations
• [x] distributions by clinical characteristics

Shiny team

• [ ] page one - plot color picker, plotly - only show certain types of cells, pick summary plots (only cd3, all, tumor ...), ripleys k - choose edge correction
• [x] page two - heatmap cluster columns/rows/neither , violin plots by clinical characteristics
• [x] page three - readme/help info
• [x] custom css

Shiny tutorials : https://shiny.rstudio.com/tutorial/written-tutorial/lesson1/ https://deanattali.com/blog/building-shiny-apps-tutorial/

[ACSoupir]

Univariate testing

• [x] Ripley's K tab to work without clinical and summary
• [ ] fix if there are only 1 cell type
• [x] plots for cumulative distribution functions for a given marker and AIC table

Multivariate Page

• [x] PCA side with side with heatmap

EDIT: Can make spatial page work without clinical and summary - currently failed due to missing data.table = F on import as well as requiring clinical (?) for the Ripley's plot? Fixed. Pushed.

[cwilso6]

• [ ] Generate a Rmarkdown report to show all plots and tables.
• [x] Fix title for PCA plot (principal is spelled wrong)
• [x] Add log and logit transformation
• [x] Change 'Modelling' section title to 'Statistical Modelling of XXX' where XXX the marker selected above
• [x] Add statement to AIC title that lower is better
• [x] Change selection titles from 'plot' to 'plot and test' for 'Choose Clinical Variable to Plot'
• [x] Change title from CDF plot to 'Cumulative Distribution Plots'
• [x] Center Model Fit underneath the CDF plot we will add a sentence of description for the model that was selected.
• [x] Add Download report button (remove download plot buttons?)
• [ ] Repeated measures are not correct eight models, need to use mix_models in GLMMadaptive
• [ ] Need to modify the code for the CDF plots to make sure they are correct for the individual markers
• [ ] Add summary statistics for given marker (Mean count, mean percent, prob == 0)

[ACSoupir]

• [ ] Update Getting Started RMD with information about modeling, PCA, transformations
• [ ] Remove other models from CDF curve
• [ ] Make dynamic modeling description below/next to CDF curve
• [ ] Make dynamic modeling description for Univariate Report
• [ ] Check interpretation for repeated measures vs non - if different check and store whether repeated measures is used and change interpretation
• [ ] NN table by plotly?