Research On Genome Assembly And Prediction Based On Deep Learning

With the continuous accumulation of biological data,people have obtained a very large amount of genomic information,relying on slower biological experimental analysis and mathematical statistics methods can not meet the needs of gene prediction,which will greatly affect the follow-up work.We need more advanced computer technologies to find the inherent meaning of sample data from a higher level.How to complete the task of gene prediction quickly and accurately has become an urgent problem.Deep learning has already been applied in many industries and has achieved good results.This paper has carried out research based on deep learning technology.The main work and achievements are as follows:?1?The whole genome assembly and genome annotation work was completed using the Citrus sinensis mitochondrial genome sequencing data.A GEC-CNN?Gene Error Correction Convolutional Neural Network?gene sequence error correction model based on convolutional neural network was proposed to correct the mitochondrial gene sequence.The complete Citrus sinensis mitochondrial genome has been submitted to the GenBank database and is officially included.The NCBI accession number is NC₀37463.This will be of great help to the molecular identification,genetic diversity and phylogenetic classification of Rutales plants.?2?This paper proposes a GP-ANN network model based on artificial neural network?Gene Prediction Artificial Neural Network?,which trains and predicts the sample gene data,and finally obtains 95.87%test accuracy on the test set.?3?This paper proposes a GP-CNN network model based on convolutional neural network?Gene Prediction Convolutional Neural Network?,which trains and predicts the sample gene data,and finally obtains 97.91%test accuracy on the test set.