Analysis of shotgun proteomics datasets requires techniques to distinguish correct peptide identifications from incorrect identifications, such as linear discriminant functions and target/decoy protein databases. We report an efficient, flexible proteomic analysis workflow pipeline that implements these techniques to control both peptide and protein false discovery rates. We demonstrate its performance by analyzing two-dimensional liquid chromatography separations of lens proteins from human, mouse, bovine, and chicken lenses. We compared the use of International Protein Index databases to UniProt databases and no-enzyme SEQUEST searches to tryptic searches. Sequences present in the International Protein Index databases allowed detection of several novel crystallins. An alternate start codon isoform of betaA4 was found in human lens. The minor crystallin gammaN was detected for the first time in bovine and chicken lenses. Chicken gammaS was identified and is the first member of the gamma-crystallin family observed in avian lenses. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1007/s12177-009-9042-6) contains supplementary material, which is available to authorized users.