Brendan Innes

If embedding names are capitalized, getEmb fails. ``` sce

Need to check on this. If not, here's the code to fix it. ``` plot_tsne(getEmb(seur,"umap")[order(getExpr(seur,"RNA")[GOI,]),], md=getExpr(seur,"RNA")[i,][order(getExpr(seur,"RNA")[GOI,])], md_title=GOI) ```

Reduce number of warnings returned to console by Shiny

2

Functions dependent on unfinished calculations are returning errors that Shiny returns as warnings to the console until the calculations they depend on are completed. These warnings aren't indicative of errors,...

Weird Shiny bugs when using seurat v1 input files.

2

One of the gene-of-interest input boxes only shows up after swapping radio buttons, and volcano plots with significant genes throw an error until swapping radio buttons for gene labels. Bugs...

See https://jef.works/blog/2019/02/17/automate-testing-of-your-R-package/https://jef.works/blog/2019/02/17/automate-testing-of-your-R-package/

Marker combinations?

2

Think about sets of genes that uniquely mark clusters (ie. CD4+ CD8-), rather than single gene markers.

See https://shiny.rstudio.com/articles/modules.html

Sonya's immunology buddies asked if genes could be viewed on a biaxial plot, a la FACS analysis. Given that each cell library is an incomplete (~10%) sample of that cell's...

Silhouette plot labels clusters by integers from 1 - # of clusters. If clusters are named, or numbered differently (Seurat counts clusters from zero because apparently they forgot they weren't...

Concerned that when selDE inputs are present, the objects used in the function are assigned to their own memory slots which would functionally double (or more) the memory footprint. Need...