Open fbastian opened 4 years ago
In GitLab by @marcrr on Jul 1, 2016, 15:27
Maybe a grey scale according to the max rank in the source? -> Always black for RNA-seq and microarray, darker or lighter grey for ESTs and in situ.
In GitLab by @fbastian on Jul 1, 2016, 16:02
Ha yes, maybe a grey scale for ESTs and in situ based on rank for this data type before norm. between data type
/max rank in the condition for this data type
In GitLab by @fbastian on Jul 1, 2016, 16:03
But it would require some extra work, maybe for now just a black or grey status based on a simple rank threshold for EST and in situ.
In GitLab by @fbastian on Jul 2, 2016, 14:19
Something like that? I'm not sure to like it, it makes the display quite complex I think. What do you think @marcrr (see screenshots below)?
(simple threshold for now, I darken scores >= 20000 with only EST and/or in situ)
Another nicer example:
In GitLab by @marcrr on Jul 2, 2016, 16:18
Ha yes, maybe a grey scale for ESTs and in situ based on rank for this data type before norm. between data type/max rank in the condition for this data type
Yes exactly!
Something like that?
Yes, I think that this is much better, thanks.
Suggestion: replace "scores are darkened" by "scores are in lightgrey"
In GitLab by @fbastian on Jul 2, 2016, 22:05
Temporary solution with basic threshold implemented, see 987b44bd8 and 1c43d7f57.
Any permanent solution should be implemented in bgee-core
, that is best suited to be responsible for determining what is a score of low confidence.
In GitLab by @fbastian on Jul 1, 2016, 14:45
For conditions with only ESTs and/or in situ data, a high rank is not likely to mean that the gene is lowly expressed, but most likely that we had no granularity enough in the data to appropriately rank the gene.
We should find a way to convey this information to the user.
Maybe put in gray the scores > 15000 with support only from ESTs and/or in situ? @marcrr