MaayanLab / harmonizome-issues

0 stars 0 forks source link

meaning of thresholdValue and standardizedValue? #22

Open moe1619 opened 1 month ago

moe1619 commented 1 month ago

Thank you so much for this tool.

Can you clarify the meaning of thresholdValue and standardizedValue when pulling data using the API? I cannot find explicit mention in the documentation of how thresholdValue and standardizedValue are created. How should I interpret thresholdValue? Should I use only genes with standardizedValue < thresholdValue?

Thank you and apologies if this is in the documentation and I missed it.

For example, using the API....

[[1]]
[[1]]$attribute
[[1]]$attribute$name
[1] "Urinary incontinence"

[[1]]$attribute$href
[1] "/api/1.0/attribute/Urinary+incontinence"

[[1]]$dataset
[[1]]$dataset$name
[1] "CTD Gene-Disease Associations"

[[1]]$dataset$href
[1] "/api/1.0/dataset/CTD+Gene-Disease+Associations"

[[1]]$associations
    gene.symbol             gene.href thresholdValue standardizedValue
1          DRD1    /api/1.0/gene/DRD1              1           1.20394

or

[[1]]
[[1]]$attribute
[[1]]$attribute$name
[1] "Urinary Stress Incontinence"

[[1]]$attribute$href
[1] "/api/1.0/attribute/Urinary+Stress+Incontinence"

[[1]]$dataset
[[1]]$dataset$name
[1] "DisGeNET Gene-Disease Associations"

[[1]]$dataset$href
[1] "/api/1.0/dataset/DisGeNET+Gene-Disease+Associations"

[[1]]$associations
   gene.symbol              gene.href thresholdValue
1         ARSA     /api/1.0/gene/ARSA              1
2         FBN2     /api/1.0/gene/FBN2              1
diamantido commented 15 hours ago

Apologies for the delayed response! It's great to hear you've found the resource useful.

Standardized values are available for some datasets where there is some measure of the strength of the association present in the original data. During processing, we perform some form of normalization and transformation to these values to create the standardized values.

When present, they can be used to compare associations within gene sets or datasets.

Threshold values are available for every association. For positive or unsigned associations, the threshold value will be 1, and for negative associations it will be -1.

I'll add this information to the documentation so it's more clear. Thank you for pointing out this gap, and hopefully this explanation is useful!