Basic Bioinformatic Studies Of Proteomic Expression Profile And Multi-modality Distribution Of PI Values In Whole Proteome

Human genome project provides us the genomeic map of human beings. In order to elucidate the genome function, human proteome project was proposed, aiming to observe the protein expression profile and their functions in different cells or tissues. Recently, the merchant instruments can identify proteins in a high throughput, and produce far more data than before. But the bioinformatics supports for further data processing lag the overload data production.The major object of this study is to establish an efficient data processing and analyzing system, which includes the following categories:Firstly, the multi-step protein identification strategy was established by using one relative high-quality, non-redundance protein database for the basic identification, and other databases for the complementary identification and novel protein mining.Secondly, the methods were established for the quality control of identified peptides or proteins. Considering, more than one peptides or proteins would match to the same identified result with the same probability. Group model was established in the peptidc and protein level. There are other special quality control methods for the PMF results: The noise data and divergence result were collected by picking out the high-frequence mass data of PMF and recalibration by Hough transform, and then were removed from GPS output result of MALDI-TOF/TOF; and the different proteins identified from single PMF result were clustered for separating the different but not similar final results identified in one PMF result.Thirdly, basing on the public methods, a new method was established by searching combined (normal + reverse database) and normal databases. This method was used in the evaluation of Q-TOF results, and was demonstrated to be more efficient than current ones.Finally, set-operation was used to combine the different proteins identified from...