Ideally something with a larger number of total compounds? I am aware of the MDL Drug Data Report database but this doesn't seem to be publicly available
The same group that produced DUD has an extended version, DUD-E
http://dude.docking.org
It has a lot more statistical power than DUD or MUV, as it has over 100 different datasets. The smaller number of targets in DUD and MUV can result in underpowered studies.
The problem with MUV might be lack of signal due to the extreme measures taken to ensure that the decoys look like the actives. In the only published paper that uses MUV that I am aware of, the tools tested all showed recovery of actives no better than random. Depending on your goals for the study that very low level of signal might not be appropriate (for example if you wish to see if certain parameter changes reduce or increase VS performance, since there is really no downward movement to be had MUV would not allow you to answer your question clearly).
- Badges
- Science topic
- Similar topics
- Bioinformatics
- Bioinformatics and Computational Biology