Workflow for in silico AS detection experiments. In the experimental proteomics study (left workflow), the (unknown) protein sample was digested using a protease enzyme. For a subset of the (unknown) initial peptide population the amino acid sequence was determined. This non-redundant peptide list was used for determining which loci were expressed (represented by a protein product) in the protein sample. The starting point for the simulations (right workflow) is a set of all annotated (TAIR) proteins encoded by the loci that were expressed in the biological sample. An initial peptide population was created by performing an in silico digestion of the set of annotated proteins. Note that the non redundant list of experimentally identified peptides is a subset of the in silico generated initial peptide population (grey dashed arrow). One thousand non-redundant peptide samples equal in size to the non-redundant list of experimentally identified peptides (thick lined ellipses in both workflows) were pooled from the initial peptide population. For each of the pooled peptide samples the number of AS events that could be confirmed with that sample was determined. Finally, the number of experimentally confirmed AS events was compared to the expected number of AS events which corresponds to the average number of AS events confirmed using the randomly generated peptide samples.