A Jordan Lab member recently asked what the units are for gene expression are for violin plots, providing the example below.
This is an issue applies beyond violin plots and includes any plot that shows feature expression. The units of feature expression may also vary between objects, and between modalities in the same object. The user should be able find out what this is without having to ask the config app user, so labeling should be added either on the plots themselves, or in the object dictionary. Whatever labelling is added will need to vary based on the object and the modality/assay the feature(s) are being added from.
The first step in addressing this issue would be to add a space in the config app where the user specifies what expression values mean in each assay. This should be a new field in the assay options card, where fields such as the display name currently exist. This could be either a short description of what numeric values mean in this assay, an extended field for providing methods used in generating the assay, or both.
Add a field to the object dictionary/dataset guide with the methods used in generating each assay, or the description of numeric values.
Modify the legends of plots to include assay specific information for expression score
Modify y-axis labels to include the short description of numeric values for the current assay
If 1) is implemented, allow the user to specify a custom axis or legend label with information on the object dictionary/dataset guide
Implementing 1) alone should be the most straightforward solution. 2) and 3) may become very difficult if the user enters features from multiple modalities.
A Jordan Lab member recently asked what the units are for gene expression are for violin plots, providing the example below.
This is an issue applies beyond violin plots and includes any plot that shows feature expression. The units of feature expression may also vary between objects, and between modalities in the same object. The user should be able find out what this is without having to ask the config app user, so labeling should be added either on the plots themselves, or in the object dictionary. Whatever labelling is added will need to vary based on the object and the modality/assay the feature(s) are being added from.
The first step in addressing this issue would be to add a space in the config app where the user specifies what expression values mean in each assay. This should be a new field in the assay options card, where fields such as the display name currently exist. This could be either a short description of what numeric values mean in this assay, an extended field for providing methods used in generating the assay, or both.
Add a field to the object dictionary/dataset guide with the methods used in generating each assay, or the description of numeric values.
Modify the legends of plots to include assay specific information for expression score
Modify y-axis labels to include the short description of numeric values for the current assay
If 1) is implemented, allow the user to specify a custom axis or legend label with information on the object dictionary/dataset guide
Implementing 1) alone should be the most straightforward solution. 2) and 3) may become very difficult if the user enters features from multiple modalities.