Data Mining of the Regularity of Compound Herbal Formulae for Pulmonary Fibrosis

ZHANG Tian-song,ZHANG Su,LI Xiu-juan,ZHANG Wei-wei,PAN Bao-feng,YANG Ke-min
DOI: https://doi.org/10.3969/j.issn.1005-5304.2011.02.014
2011-01-01
Abstract:Objective To study the regularity of compound herbal formulae for pulmonary fibrosis.Methods Compound herbal formulae published on journals for the treatment of pulmonary fibrosis were searched and collected.A database was established for data mining using frequency count,cluster analysis and association rule analysis.Regularity of key herbs and a pair(group) of herbs was summarized.Results There were 64 prescriptions(totaling 114 herbs and 584 counts of frequency) in the compound herbal formulae for pulmonary fibrosis.The herbs which were used for more than 5 times included 36 herbs(totaling 431 counts of frequency).They were cluster analyzed and classified into 6 clusters including tonic,hemorheologic agent,apophlegmatisant,drugs used to disperse and lower the qi of the lung,heat-clearing drugs and antiasthmatic.The rules in pair of herbs were 19,and rules in group of herbs were 25 by association analysis.The rules consisted mostly of compatibility of qi-reinforcing drugs and hemorheologic agent.Conclusions It was feasible that study the regularity of Chinese compound prescription for the treatment of pulmonary fibrosis by data mining.
What problem does this paper attempt to address?