The Research Of Gene Chip Recognition System

As the Human Genome Project proposed and implemented and the rapid development ofelectronic information technology, an increasing number of bioinformatics technologies havebeen widely applied. Particularly the gene chips technology which due to its powerfulgenomic information analysis capabilities, has been used in many areas of biological science,and becomes the focus of many research institutions. In order to realize the gene chip samplepoint signal fast, accurate and efficient testing and analysis, this paper mainly studiesdesigned a set of complete gene chip recognition system, hope to promote gene chiptechnology further perfect and application in more fields. The main work content isdescribed below.1. Study design out more scientific and efficient gene-chip identification system of generalprocess. With reference to the literature and past gene chip recognition research results inthis paper, the regular gene chip recognition system is improved, and particularly in the genechip image preprocessing stage, joined the image enhancement steps, specially for theimprovement gene chip image grey range, gene sample point and the background imagecontrast is not big problems for the follow-up of the image processing to provide the help.2. There is a certain amount of noise, bright spots and interference problems for most ofthe gene chip image. In gene chip image denoising stage, this article through the simulation,choose the filtering effect good adaptive median filtering algorithm. In the proposedalgorithm to determine the size of the dynamic filter window way, handle existing complexgene chip image noise effectively already filter out the noise, and can completely retainedsample points and the edge of gene details characteristics3. The gene chip recognition system image grid orientation process is crucial. This paperwith the improved projection algorithm to gene chip image grid orientation, this algorithmbased on arithmetic of classic join the power spectrum calculation steps. Through thesimulation results verify the algorithm can accurately calculate gene sample point intervalvalues, which can get good grid orientation effect.4. For the more accurate extract samples of the characteristic parameters of the genes atnumerical, calculates the genes of threshold image sample point is very important. Throughthe simulation of threshold segmentation algorithm, by using the least wrong threshold segmentation algorithm is the threshold. This algorithm is widely applied to the imageprocessing algorithms, not ever used in gene chip image processing, but from the simulationresults can be seen in the principle of the algorithm is simple, easy to realize, processing andbackground contrast of poor genetic sample point effct is obvious,5. At the end of the article, this paper application design gene chip recognition system,from the laboratory to make viruses characteristics testing gene chip image processing. Andgene chip to identify the system’s function and effect is evaluated. The result shows that thesystem can evaluate the success of the different types of gene virus sample point todistinguish the accuracy is higher. For the same virus different concentrations of the sampleclassification, the accuracy is not high, also need to work hard to make up this deficiencyafter, further improve.