Detection Algorithm For Autosomal Multiple Abnormalities

Chromosomal disease is a disease that causes changes in the chromosome multiples of the fetus or changes in the genetic structure of the fetus during pregnancy.With the discovery of fetal free DNA in maternal plasma,we can obtain fetal gene information by sampling maternal blood before production,and then obtain DNA sequence by gene sequencing.The first task of this paper is to detect the detection of autosomal abnormalities,that is,to detect whether 22 pairs of human autosomal have increased or decreased.The first algorithm is based on the whole chromosome detection algorithm.The algorithm consists of two modules,data preprocessing module and decision module.Data preprocessing module uses 12 processes to process the reference group sample and the sample to be tested,and calculates the Z value of each chromosome content in the sample.The decision module uses statistical test method to select the threshold of decision and define the boundary of grey area.A decision tree algorithm for autosomal classification is proposed.Then,180 real samples were tested by this algorithm to determine whether the samples contained autosomal abnormalities such as trisomy 21 syndrome.The detection rate of positive samples is 100% and the total correct rate is about 93%.The second algorithm is based on segmented chromosome detection algorithm.The core idea is to calculate the Z value of each chromosome in each sample as an important basis for judging whether the sample is an autosomal multiple abnormal disease.The combination of the two algorithms can detect multiple abnormalities of autosomal chromosomes.The second part is an algorithm based on double reference group to detect sex chromosome abnormalities,that is,to detect whether the number of human two sex chromosomes increases or decreases.The number of sex chromosomes of men and women is different,but the existing algorithms do not pay attention to the effect of the number of X and Y chromosomes of male and female fetuses on the results.Therefore,this paper proposes a new detection algorithm based on this point.The algorithm consists of two modules,data preprocessing module and decision module.Data preprocessing module is the same as autosomal detection algorithm.In the decision-making module,a Double-Reference group was selected,and pregnant women with normal fetuses were divided into two groups according to male and female fetuses.For pregnant women with male fetuses,the data composition includes the XX chromosome of the mother and the XY chromosome of the child.For pregnant women with a female fetus,the data composition includes the XX chromosome of the mother and the XX chromosome of the child.Then the decision is made according to the classification decision tree algorithm of the algorithm.When deciding whether the sample is negative or positive,the first step is to judge whether the sample is male or female.The sex of the fetus was confirmed by comparing the Y chromosome of the sample with the reference group of the female fetus.Then we compared the X chromosome of the female fetus with that of the female fetus reference group to determine whether there were polysomic or monosomic abnormalities in the X chromosome.Then for male fetuses,we not only compare them with the X chromosome of the male fetal reference group to determine whether there are polysomic or monosomic abnormalities of X chromosome.Moreover,the Y chromosome of the male fetus reference group should be compared to determine whether it has Y chromosome polysomy or monosomy abnormalities.Finally,the karyotype classification was determined.Finally,the positive sample detection rate of the algorithm is 100%,and the total correct rate is about 91%.The last work is to annotate the results of detection of abnormal mutations in nonchromoso-mal multiples.For example,single gene mutation disease,microdeletion microduplication disease and so on,and the possible drug use of the mutation is given.In this paper,we collate the database of some common cancer gene mutations and targeted drugs,and then annotate the obtained gene mutations to obtain the information of the mutated genes,bases,amino acids and so on.Then seven types of mutations were completed.Amino acid mutation,exon mutation,gene amplification,non-frameshift insertion mutation,gene fusion,gene deletion and gene homozygous deletion are the seven types of mutations.We can link these seven types of mutations to the drug library and compare the potential targeted drugs for mutation of this type of cancer gene.