Searching association rules of traditional Chinese medicine on Ligusticum wallichii by text mining

Much useful information on Ligusticum wallichii (LW) could be obtained from published literature by text mining technique. In this study, the data set on LW was downloaded from Chinese BioMedical literature database (SinoMed). Then, association rules among diseases, traditional Chinese medicine (TCM) syndromes, formulae and herbs on LW were investigated by text mining technique. These rules include TCM syndromes to diseases, formulae and combinational herbs, respectively. Diseases related with formulae including LW were mined out by executing data slicing algorithm. Finally, the results were visually demonstrated with Cytoscape 2.8 software. The main features from the mining data were: (1) LW was frequently used in treating cerebral infraction; (2) Blood stasis due to Qi deficiency was the main syndrome in TCM clinical practice; (3) Angelica sinensis was the first herb to combine with LW according to co-occurrent frequency; (4) Associated with LW, networks of TCM syndromes-diseases, formulae-diseases, TCM syndromes-formulae, and TCM syndromes-combinational herbs were constructed. These associated networks represented a holistic thinking of Chinese medicinal therapy, which might embody association rules among diseases, syndromes, formulae and herbs on LW.