| |
| Single molecular functional network construction and analysis of disease is very useful to identify novel and potential targets for prognosis and therapy. This paper integrated an infer method based on linear programming and decomposition procedure with function analysis using Kappa statistics and fuzzy heuristic cluster (DAVID). We first identified the significant molecule ABCC3, then constructed ABCC3 up- and down-stream network by infer and further data-mined the main ABCC3 modules including transporter activity, splice variant, sequence variant, cell fraction, transmembrane, catalytic activity and ATP-binding from 25 lung adenocarcinoma and 25 human normal adjacent tissues in the same GEO Dataset GSE7670. Our infer ABCC3 network result showed the different gene rate of lung adenocarcinoma as 54% (26/48) compared with the control considering activation and inhibition relationship. The different active genes in lung adenocarcinoma include ASPM, GCNT3, SPP1, SRD5A1_2, CRABP2, HIST1H4J, MKI67, PYCR1 and the different inhibitory genes include BIRC5, COL1A1_2, GINS2, GREM1_2, HIST1H4J, HLXB9, MELK, MMP11, SPINK1, TOP2A_2, ASPM, COL3A1, HMGB3, HMMR, MMP12, SRD5A1_1, SRD5A1_2, TOX3_3. Our integrative analysis showed the positive results of ABCC3 transporter activity and cell fraction through the net numbers of activation minus inhibition compared with the control and predicted the increases of these modules in lung adenocarcinoma, whereas the negative results of ABCC3 ATP-binding, transmembrane, catalytic activity, splice variant and sequence variant, and deduced the decreases of these modules in lung adenocarcinoma. |
|
| Keywords:ABCC3; human lung adenocarcinoma; network construction and analysis; integrative biocomputation |
|
