Design And Implementation Of Multichannel Bioassay Instrument Based On FPGA

Recently,bioanalytical instruments are constantly being improved.Microbial detection has also been gradually transferred from routine detection to molecular biology.Most of them are detected by DNA.Nanopore sequencing technology is a new generation of DNA detection technology.Therefore,the use of nanopore to detect microorganisms has become the cutting-edge technology of today.This paper mainly uses nanopore sequencing technology,using FPGA and other hardware as the main body of the instrument,using intelligent algorithms to classify and identify DNA types.The main research contents are as follows:First,for the phenomenon that the original current of DNA could not be obtained,the data of MinION equipment was obtained from the European nucleotide archives,and the event information was reconstructed into the original signal data,and the mif file was made by using MATLAB.The advantages and disadvantages of the three digital circuit frequency synthesis techniques are analyzed,and the direct digital frequency synthesis technology(DDS)is selected.A multi-channel analog data generation device is designed and used as the experimental data source of this thesis..Secondly,a high-speed analog-to-digital conversion circuit is designed using ADC9226 chip,and FPGA programming is used to collect the data generated by the data source at high frequency.At the same time,a UDP communication module without MAC IP core is written using Gigabit Ethernet module.The data collected is forwarded to the host computer,and the experiment proves that the working speed can reach Gigabit,which can meet the real-time data collection.Then,the principle of biological nanopore sequencing was analyzed,and the two root causes of error were analyzed.The traditional base recognition algorithm and feature extraction method are introduced.On this basis,a new method for feature extraction of DNA detection is proposed.The Hjorth parameter in EEG domain is used as the feature extraction method of DNA original signal data,and a new DNA species recognition classifier design is constructed.Using this classifier can greatly reduce computational complexity and decision time.Finally,a variety of artificial intelligence classifiers were described,and the support vector machine classifier and random forest classifier were used to classify and verify the 180 sets of E.coli and Pseudomonas fluorescens sequencing data by python.At the same time,the grid search,genetic algorithm and tree Parzen evaluator algorithm(TPE)were used to optimize the hyperparameters of the two classifiers.The final optimized classifier was used to plot the ROC curve to compare its performance.Finally,the genetic algorithm was found.The random forest algorithm optimized by TPE algorithm has a good effect on the classification of the two.