Abstract
In general, a large amount of DNA mutations accumulating at oncogenes and tumor-suppress genes are the main cause for cancer. Since the differential expressions of these oncogenes and tumor suppress genes are primarily found in the signal transduction pathways (STPs), the signal transduction abilities or signal transductivities of the pathways in cancer cells need to be investigated systematically by their microarray data and then compared with those in non-cancer cells. Based on systems analysis through the samples of microarray data in leukemic cancer cells and non-leukemic normal cells, we found that the transductivities of the transcription factor (TF) STATs are enhanced in leukemic cells while the transductivities of the other TFs including p53, C/EBPa, CSLs etc are attenuated in leukemic cells. Further, the transductivity sensitivities of the proteins in coupling STPs of leukemic and non-leukemia normal cells also reveal their roles in the carcinogenesis process of leukemia. Based on transductivity analysis, we trace back the main cause of dysfunctions of TFs, detect mutant proteins such as DVL, PSE2 and ERK, and discuss the role of crosstalks between multiple STPs in the leukemogenic process. Most importantly, we provide lists of proteins as signatures for the identification of the progression from myelodysplastic syndrome (MDS) to acute myeloid leukemia (AML) and specific subtypes based on transductivity analysis. Since TFs at the downstream of STPs simultaneously play an important role in the corresponding gene regulations in nucleus for cellular functions, the network robustness of gene regulatory network (GRN) consisting of these TFs in leukemic cells is also estimated by the samples of microarray data and compared with those in non-leukemia normal cell to investigate the systematic property of GRN in leukmogenic process. It has been demonstrated that leukemic cells with larger network robustness, especially at AML, can tolerate more intrinsic genetic mutations to provide a diversity of heterogeneities for evolution than non-leukemia normal cell.