The exploding growth of the biomedical literature presents many challenges for biological researchers. One such challenge is from the use of a great deal of abbreviations. Extracting abbreviations and their definitions accurately is very helpful to biologists and also facilitates biomedical text analysis. Among existing approaches, text alignment algorithms are simple, effective and require no training data. However, state of the art alignment algorithms could not identify the definitions of irregular abbreviations (e.g., <CNS1, cyclophilin seven suppressor>). We propose an algorithm analogous to pairwise sequence alignment, in which it is given a penalty score if there are two unmatched characters separately from the abbreviation and definition, and in this way some irregular abbreviations are found.