Why are there only two choices for the confidence (>80% or ANY) in queries on gene differential expression?
The differential expression analyses were all run at the reporter level, so confidence values are attached to each reporter. A gene might be represented by more than one reporter, so one doesn’t have a confidence value automatically attached to a gene. In the current release, when we say that a gene has confidence >80%, it means that it has at least one reporter with that confidence and there are no conflicts (i.e. no 2 significantly diff. expressed reporters where one is up and the other is down). The main issue that prevents from filtering by different confidence thresholds in differential expression queries has to do with difficulties in linking these to gene fold-changes (which are also obtained as summary statistics of reporters fold-changes) in a meaningful and consistent way.
