Is there a way of making a random selection of compounds from my database from the command line? Can it be done by specifying the number of compounds rather than the percentage?
This cannot be done directly. However, the utilities sdsubset and maesubset support arbitrary selection of compounds. You could then write a small script or program to generate random numbers in the desired range, and write them to a text file, which can then be used as input to these utilities with the -n option. The output from the utilities is the desired random subset.
You can make a random selection of compounds in Maestro.