I am working on creating a knowledge graph using data from the Open Targets Platform (Open Targets Platform). I have a specific question regarding the creation of edges between nodes. Could you clarify the meaning of the different association files available on the Open Targets Data Download Platform? I am including a screenshot of the datasets related to associations for reference. Your guidance on how these associations are defined and their significance would be greatly appreciated.
Dear Shailesh,
thank you for your question.
You can find all your answers on our dedicated documentation page, including definition and calculation of association score, explanation for the ontological selection of evidence (direct or indirect associations) and difference between overall score and score by data source or data type. Platform data types are: genetics associations, known drugs, pathways & systems biology, RNA expression, somatic mutations, text mining, animal models - as it will be clear from the files you are accessing. File schemas are also accessible on the right side of the data downloads page.
Please do let me know if something is not clear enough, and good luck with your project!
Best wishes,
Annalisa
Thank you for your response @Annalisa_Buniello ,
After looking documentation page i get to know different data type and data source, i was looking at association files present at open target data download page, there is two different type of association given like direct and indirect, what is meaning of direct and indirect association, on documentation page its mention that some disease have hierarchy of diseases that present in indirect association file and disease have no hierarchy are present in direct association, so it is like direct association is subset of indirect association? so i can use indirect association file only to make edges between nodes? I am little confused.
Also there is a evidenceCount column present in association file what this column value signify.
- Indirect Association: Includes evidence that links the target to related diseases within the disease’s ontological tree (e.g., subtypes or broader categories).
Hello! This is not quite correct — indirect evidence only includes evidence associating targets to child terms of the disease in the ontological tree.
For example if you were to look at the evidence for “Cancer”, you would find indirect evidence for associating targets with “Breast cancer”, but if you look at the evidence for “Breast cancer”, you will not see evidence associating target to “Cancer”.
Thank you very much! This clarification is extremely helpful!