The Design And Implementation Of Data Analysis Method Of Digital PCR Fluorescence Detector

The concept of digital PCR was explicitly proposed by American scientists Vogelstein and Kinzler in 1999,The microsamples were graded and diluted with a large scale until the number of tested molecules in each sample was no more than one,and all samples were amplified by PCR under the same conditions.The sample contains a target molecule that magnifies millions of times and produces very strong fluorescent signals.Then the sample of the amplified sample was counted and the copy number of the target nucleic acid contained in the original sample was obtained at last.digital PCR usually can be used for the prediction of disease and the diagnosis,it can also be used to guide people to supervise and control the disease,which guides people to safe medication,help doctors customize personalized treatment,convenient and effective treatment of disease.The digital PCR instrument in this paper is composed of droplet generator,thermal circulator,fluorescence detector and a set of upper computer.DNA samples under test is divided into tens of thousands of droplet after droplet generation system,the DNA molecules of target are amplified by the heat cycle,finally determining whether each droplet contains the target sequence(PCR extension or genes)by fluorescence detection system.To count,the fluorescence intensity of target DNA was detected by two fluorescent dyes,FAM and VIC.Liquid droplets with high fluorescence intensity are considered positive,and little or no fluorescence drops are considered negative.The construction of hardware platform is introduced in detai.The focus includes the principle of droplet formation?the design structure and data acquisition module of optical path.The hardware platform mainly completes the function of data collection.The data is collected mainly through PMT,and then USB communicates with STM32 and FPGA.The upper computer software mainly completes the data processing functions.Mainly include software framework design,partial module description,1D graphic design and implementation,2D graphic design and implementation,threshold line and rigging tools,etc.By comparing the various algorithms Using the k-means clustering algorithm in machine learning to divide the threshold of negative and positive droplet.determining the threshold value and number and other parameters of negative and positive droplet is to achieve the purpose of genetic testing.