Methods and apparatus are provided for the identification of one or more candidate chemical formulas from mass spectrometry data corresponding to an unidentified chemical compound. By restricting the generation of candidate formulas to those having repeating units and/or end units with specified limitations, the methods and apparatus may more efficiently iteratively search for a chemical formula having matching mass spectrometry output within a threshold tolerance. In another aspect, methods and apparatus are provided for the identification of one or more candidate chemical formulas from mass spectrometry data based at least in part upon neutral loss.