Accumulation of genomic data is critical to our understanding of the overall functionality of a species. To facilitate designing microarray experiments for poorly-studied yeast species, we developed this system to recommend the most informative experiments based on available Saccharomyces cerevisiae data. The recommendation of the experiments is based on the following three factors:

1. Overall accuracy of an experiment in capturing protein functions.
2. Overlap between different experiments represented by their conditional mutual information.
3. Existing list of experiments in the poorly-studied yeast species.

Here, α∈[0,1] represents the trade-off between accuracy and overlap of the datasets. When α=1, we recommend experiments only based on their individual accuracy; and when α=0, we recommend experiments only based on their similarity from existing datasets. An α between 0.3 and 0.9 is usually preferred, depending on the purpose of your experiments (whether to achieve accuracy in integration or achieving accuracy in function prediction).