RESUMO
Objective Based on data mining technology, we discussed the correlation of different gene mutation, fusion or amplification in advanced lung cancer patients. Methods 630 cases in our hospital were collected and the drug database was established by Excel 2016. Frequency analysis and association rules were used for statistical analysis, and statistical analysis was performed on SPSS22.0 and SPSSModerler statistical software.Results There were 26 high frequency genes, such as EGFR-19 mutation, ALK fusion, RET mutation, EGFRL858 R-21 mutation, ALK mutation, and the first 6 cases of T790M mutation. The first 26 were NRAS-G12 D mutation and BRAF-G466 V mutation, MYC-R450 W mutation and CYP2D6 mutation, GATA3 M423 fs mutation and ESR1 mutation, abrupt mutation, mutation and sudden process. The variable classification values are associated with each other. The minimum value of gene association analysis was 50％ and support was 10％. After Apriori module analysis, EGFR-19 mutation and 14 genes were not found (confidence degree 46.667％) , EGFR-19 mutation and T790 M mutation (confidence degree 40％) , EGFR-19 mutation and TP53 mutation (confidence degree 13.333％) , ROS1 fusion and MET amplification (confidence 47.619％) , MET amplification and ROS1 melting (confidence level 50％) , T790 M mutation and EGFR amplification (confidence level 57.895％) , EGFR-19 mutation and EGFR amplification (confidence 42.105％). Cluster analysis BRAF-G466 V and NRAS-G12D, MYC-R450 W and CYP2 D6, GATA3 M423 fs and SRC, PIK3 CA amplification and PIK3CA, Pten and EGFEL861 Q-21, KRAS G12A and blending. Conclusion There are correlations between different genes and mutation, fusion, lack and amplification of lung cancer.