@padmashris Could you add source to the previously mentioned functions?

They both call verify_factor

Thanks!

@DJT333 change the list of data to be named using the tester object with names to enter the mk_mega_tbl function

-Capitalize the final group names

is there something that's missing here? I don't fully understand what the function is doing

lines 62-67 from the 01_assign_metadata_make_initial_object.R file:

#order them
invisible(lapply(objects, function(name) {
        df <- get(name)  # Get the data frame from the environment
        modified_df <- sort_dataframe(df)  # Apply the function to the data frame
        assign(name, modified_df, envir = .GlobalEnv)  # Update the data frame in the environment
}))

Add to Supplemental folder

Add as a separate script in CreateObjectsAndClusters

Combined 01_ and 02_createObjectwithmetadata

-Remove pseudogenes & model genes

F2stats is the first attempt to calculate the proportions for F2

Steps in my head:

1. calculate counts with calc_counts
2. split by rat_group then calculate proportions
3. find z score from the proportions with homecage as control
4. run 2 way anova on the z scored data

Uploaded F1 script and add_STreads.R

@DJT333 can you make changes to line 22: use source() and not have a dependency on the output of another script

@ksavell can you push the gaba_v2 file that you modified once again?

If the cluster name has an underscore, there is an issue with prep_upset and make_upset.

~Line 17 within prep_upset, involves string split and underscores!

Added code from Kareem when he manually plotted these into the start of the function file.

general plots have been color coded with the hex codes from the Formatting document.

need to pick colors to split by groups, ratID, and sex.

Please help!!

I define clusters to keep in the subset as clust_tokeep. When I run line 57 I get

Cannot find the following identities in the object: 0123456781213141516171920212429

I think this means it's not separating the clust_tokeep into different idents.

calc_proportions.R

When it calculates the proportion, it's for the total of all factors.

For example, if we are using it to calculate cluster populations for each rat/population, we need it to calculate the proportion relative to the nuclei collected from each rat's population.

bla bla

In figure2.R line 66, I expect all clusters to be ran for glut since there are > 100 nuclei per cluster in each cluster.

         CTL6 CTL6b ITL23 ITL56 NPL56 PTL5

Active 334 6 525 737 26 233
Homecage 2817 295 2824 2913 464 1271
Non-active 1121 100 1306 1262 180 598

Table:
Non-active Homecage
ITL56 737 2913
ITL23 525 2824
CTL6 334 2817
PTL5 233 1271

However, I think it's still using Active for the inclusion criteria despite what is in the comp_vect spot

Trying to automate this to produce nCount_RNA and nFeature_RNA for glut and gaba:

-set the plot to be a certain size. If testing between glut/gaba there is a difference in cluster name length, and the violins get squished. In the plot it's 1.5 x 1.5 inches.
-No labels, was adding them back as correct font, size, and color.

I was examining the resulting objects after line 58 and noticed that the gene names under the RNA counts have the cartridge identity as a prefix

Example: c2.X0610005C13Rik

Found this under:
data_C2@assays[["RNA"]]@CountS@Dimnames[[1]]

(Adding here to keep organized)

Can there be another tab or option for 'cluster specific plots' where there is a dropdown for which cluster you want to look at?

Right now when it groups by group/sex/rat it collapses by cluster

Giving this a try :)

I tweaked a few things, but here are things I don't know how to change immediately:

Can the correlation plot between 2 genes be worked in? Perhaps it has to be a different tab? The first tab could be 'Single gene analysis' instead of 'GABA dataset' on the output section, then another tab for 'Multiple gene analysis'. This is me assuming multiple tabs are an option

Can all features be italic? This will conform with gene naming standards.

I noticed on the 'split by: group' feature plots, the i's are capitalized in negative and positive.

Can you make the colors specific to the hex codes in the formatting file? This might be a larger issue that we need to mod the base seurat objects (along with 'renaming' the clusters from the numbers to the cluster names)

PS: I like the hover color on the drop downs!