A micro-droplet data analysis method and system

Abstract

The invention discloses a method and system for analyzing micro-droplet data, which can efficiently cluster and analyze the micro-droplet data, and improve the accuracy and stability of the analysis results. The method includes: Step S1: Obtain m sample data from the micro-droplet sample, m>>3; Step S2: Use the K-means algorithm to preliminarily classify the m sample data to obtain Vic signal clusters, fam signal clusters and Noise signal cluster; step S3: calculate the ratio result of the fam signal cluster and the Vic signal cluster, when the ratio result belongs to the threshold range, perform step S5, otherwise, perform step S4; step S4: pass the K mean value algorithm to the Vic signal cluster Optimizing correction, or optimizing and correcting fam signal clusters by DBSCAN algorithm, return to step S3; step S5: output the cluster distribution diagram of Vic signal clusters, fam signal clusters and noise signal clusters, and obtain the data analysis results of micro-droplet samples. The system includes the method proposed in the above technical solution.

Description

technical field [0001] The invention relates to the technical field of biological reaction numerical simulation calculations, in particular to a micro-droplet data analysis method and system. Background technique [0002] Droplet digital PCR technology is a new type of nucleic acid detection method. Its principle is to evenly distribute each sample reaction solution into a large number of emulsion-wrapped micro-droplets, and perform PCR amplification independently in each micro-droplet. Then, the fluorescent signals of the droplets are detected one by one by a method similar to flow cytometry, the number of positive reaction wells and negative reaction wells is determined, and the number of nucleic acid copies in the sample is calculated by the statistical method Poisson distribution. [0003] In the process of analyzing the micro-droplet data, the core step is to cluster the micro-droplets, and then output the analysis results of the micro-droplet data according to the clus...

AI Technical Summary

Problems solved by technology

The existing micro-droplet data analysis methods usually use a two-dimensional scatter diagram that can display the parameters of two measurement channels, and perform cluster analysis on the data in the two-dimensional scatter diagram by manually setting a gate. With the rapid development, the data volume of detection parameters has also increased exponentially, and the traditional analysis method of manually setting gates is no longer able to meet the huge micro-droplet data analysis business.

Images