aertslab / scope Goto Github PK

try e.g. drawing 8

to support regulon view and comparison view, a nice feature would be that this function would take into account:

threshold values (per each regulon) to filter out cells passing (coloured) or not (black)
intesity flag: true/false to take into accout auc values scaling (true) or just full rgb (false)

currently features needs to be selected in R G B order

extend getCellColorByFeatures by new arguments:

coordinatesID
ageingAnnotationID
add possibility to put clusters next to genes and regulons

to colour complete cluster

somewhere in visible place so that user knows that backend is doings something.....

A very nice feature that can enrich the investigation of the data, is for points to be plotted based on their expression level, with the point on the top (plotted last) being the higher expressing cells.

The question is, whether this is easily implementable given that most of the frontend is working off of indices rather than the cellIDs?

Thoughts @kreftl ?

The Regulon AUC values plot should colour all points regardless of the threshold

currently the R, RG, or RGB colouring is supported
alternatively user needs also to have access to G, B, GB and RB colourings
so for query e.g. {
0: {type: "gene", value: ""}
1: {type: "gene", value: "VGlut"}
2: {type: "gene", value: ""}
}
the getCellColorByFeatures should return colors array iso null

Include clusterings (i.e. all clusters) and individual clusters from each clustering

If you type in the query box, but tab to the next one without clicking a gene or pressing return on it, it won't have any effect

Make installation as simple as possible

for empty queries, e.g.
{
0: {type: "gene", value: ""}
1: {type: "gene", value: ""}
2: {type: "gene", value: ""}
}
getCellColorByFeatures should return cells coloured in base color (black)

normall, add, multiply or screen ?

for the lasso selected cells

get cluster mapping table for different loom files (if exists)

to obtain data like this there is need for backend call

The plot should be updated when the log or CPM normalisations are toggles on and off to update the values shown. Will need to request features from the server again

Refactor this function to send less data in the initial burst an allow the frontend to request it as and when needed

On the dnd tab, when you drag an annotation onto a space when you have genes already selected, you first see colours on the wrong cells

We have to deal and warn the user when loom files are invalid/corrupted or valid but don't have the extended version of the loom (e.g.: RegulonsAUC not present or any other attributes not present). How should we do this?

to support comparing tSNEs

Regulons don't use CPM or log normalisations and don't need to be redrawn

CPM Normalisation should be off by default.
Both CPM and Log2 normalisation settings should be shown on each tab - Possibly under a 'Gene expression' subheading

accordingly: app, header, footer, leftsidebar, rightsidebar, content for each tab

Separated by feature type
Possibly show a minimum of X results

When a loom file is selected, we should disable non relevant tabs

the more times you select cells the slower it gets

See c088d8d

running servers on windows for a longer period of time results in memory error

Add function for retrieving markers from gene sets, clusters and regulons.