Gender discrimination method based on chip detection of DNA data

Abstract

The invention discloses a gender discrimination method based on chip detection of DNA data. The gender discrimination method includes the steps: selecting S sites on all sex-related X and Y chromosomes and reading P samples to obtain a P*S size matrix, and screening M sites with large discrimination; establishing a logistic regression model for each site among the M sites, wherein the independentvariables are DNA detection red fluorescence and DNA detection green fluorescence values and the dependent variable value is 0 or 1, representing male and female, and obtaining M regression models andsaving the M regression models; and when predicting a new sample, through the logistic regression models of M sites, obtaining M prediction values between 0 and 1, and calculating the final probability according to the probability interpretation of the logistic regression prediction values. The gender discrimination method based on chip detection of DNA data combines the X and Y chromosomes, andis fully quantified by the method of machine learning, and the gender discrimination result is a value between 0 to 1, thus being able to be compatible with various data cases and making richer discrimination for gender.

Description

technical field [0001] The invention relates to the technical field of DNA detection, in particular to a gender determination method based on chip detection of DNA data. Background technique [0002] The gender judgment based on the chip detection DNA data is generally based on the typing of the X and Y chromosomes in the DNA data. There are two common methods: 1) Only through the typing data of the X chromosome, in most cases, women have 2 X chromosomes, the typing data has two types: homozygous and heterozygous, and men have 1 X chromosome, There is only one type of homozygosity in the typing data. Generally, the gender can be confirmed by counting the homozygous or heterozygous rate of the X chromosome typing data of a certain sample; 2) only through the typing data of the Y chromosome. In most cases, females If there is no Y chromosome, there should be no Y chromosome data. Males have 1 Y chromosome, and the typing data has 1 type of homozygosity. Generally, the gender ...

AI Technical Summary

Problems solved by technology

[0003] The above two methods can judge most cases, but many abnormal cases cannot be identified

For example, some people have X chromosome or Y chromosome abnormalities, which will affect the accuracy of judgment. In addition, the above two methods are based on traditional statistics, and general judgments based on experience cannot make more accurate predictions.

Images