Missing RNA expression data in the overall score (by disease)

could no see any data link to RNA expression. Is it a ‘bug’ ?


Dear Nicolas,

The result you are observing is the expected behaviour although there is a long explanation why this happens.

We suspect your querying a high-level term in the cancer branch. When looking at a disease evidence for all the targets it’s propagated up in the ontology. For example, if you are looking at cancer, some of the blue cells might be the result of propagating information up in the ontology from descendant terms like breast cancer. More info in: Target - disease associations - Open Targets Platform Documentation

However, we observed RNA expression evidence can be very promiscuous and sometimes pollute the associations of higher level terms. For this reason, we blocked the the propagation of evidence for this data type in particular. If visiting cancer, and there is no direct evidence linking the term cancer with specific targets, the cell will be white.

Apologies because this behaviour was not made very clear in our documentation. We have added a disclaimer in the respective section.



Hi David,
many tks for your answer, and I ‘think’ I understand :slight_smile: indeed, it was rather large ontology query: ‘neoplasm’ :wink:

1 Like

I can understand both of you, but white still will make our users confusing, I think. Do we have any better proposal for such situation?


The next plot it’s from data 18 months ago, but the same principles apply. At that time we had a large number of differentially expressed genes informing about 186 unique diseases (colour gradient represents the number of associated diseases)

I’m very open to listening to other proposals on how we could improve the current ad-hoc rule for expression atlas evidence - Target - disease associations - Open Targets Platform Documentation

@ochoa This differential expression result is only available for disease-based look-up, not target-based look-up, correct? Whether this differential expression data is downloadble for given targets among 186 diseases?