Dear goseq developers!
I was wondering whether there are any plans to implement the conditional test from the GOstats package also in goseq? That test was pretty handy to not get all these very general GO terms being significant.
cheers, jo
It's more up to Nadia Davidson, who is the maintainer of goseq, than to me.
However, we tend to spend time on things that we like to use ourselves in our own collaborations, and I and my collaborators quite like to include the more general GO terms in the results table. (One of my collaborators has even asked to get rid of the less general GO terms, so that only a big picture would emerge.) If you could point us to a compelling example where the conditional analysis was definitely an improvement, maybe you could change our minds. It's not a trivial feature to add though.
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
version 2.3.6
dear Gordon!
Thanks for your answer. There is no urgent need to such an option in goseq. On the other hand, the idea to remove the less general terms is also quite interesting. In the end it would just be nice to reduce the number of overlapping and sometimes redundant terms in some way (remove the less general, or the more general, or group the terms afterwards), but I know that's not straight forward to implement.
thanks, jo