KR20100060614A

KR20100060614A - Fast automated analysis method of dna microarray

Info

Publication number: KR20100060614A
Application number: KR1020080119277A
Authority: KR
Inventors: 이진호
Original assignee: 이진호
Priority date: 2008-11-28
Filing date: 2008-11-28
Publication date: 2010-06-07

Abstract

PURPOSE: An analysis of DNA microarray is provided to improve analysis speed and without error even when array lattice is declined and to maintain lattice form. CONSTITUTION: A completely automatic analysis of DNA microarray comprises: a first step of extracting a boundary component using canny edge detection from a DNA microarray; a second step of regarding weak boundary component as noise to obtain image removing the weak boundary compound; a step of obtaining image in which open figure is removed; and a step of repairing damaged spot to obtain DNA probe.

Description

High speed automated analysis of DNA microarrays. {Fast automated analysis method of DNA microarray}

The present invention relates to a DNA microarray, and more particularly, to a DNA microarray analysis method which improves the noise ratio and speed of a fully automated analysis using Hough transform and block seperation techniques.

DNA microarrays are primarily aimed at finding out how much DNA sequences are expressed in a particular tissue. The experimenter's biochemically produced DNA arrays are called probes, and microarrays are made of 20,000 to 40,000 round grooves in a lattice shape on a 2cm x 4cm glass plate, and the probes are stained with phosphors. will be. The stronger the expression level of the DNA, the stronger the color. The analysis of microarray means that the position of each probe is identified to quantify the intensity array.

There are many ways to analyze DNA microarrays. At present, the most commonly used method is to find out the size and spacing of probe spots manually, but this method is cumbersome and requires a new spacing every time multiple image specifications are different.

In the case of fully automatic microarray analysis without receiving any information from the user, the well-known Canny edge detection is used to find the edge of the probe spot and then remove the component that is considered noise. At this time, normal spots that are not noise disappear as well. If there is no spot in a space over a certain size, it is estimated that there is a spot at a horizontally or vertically parallel position from the surrounding spot and extracts color.

However, this method is slow because it processes a lot of spots sequentially, and it is assumed that there is a normal spot at the horizontal and vertical parallel positions from the surrounding spots. There is a problem. Because of these problems, they are rarely used in practice.

The present invention is to solve the problems of the prior art, the purpose of the accurate analysis by reducing the misrecognition of the spot position when the entire microarray image is tilted.

It also aims to improve overall speed by reducing the time wasted by sequentially solving all probes one by one in processing large microarray images.

DNA microarray analysis method according to the present invention for solving the above problems is a first step of extracting the boundary component from the DNA microarray image using canny edge detection; A second step of recognizing the weak boundary component as noise and removing the removed image from the boundary component; A third step of obtaining an image from which the open figure is removed from the image obtained in step 2; The fourth step of obtaining the degree of DNA probe staining by recovering the normal spot lost in the image obtained in the third step using Block seperation and Hough transform.

As described above, according to the present invention, the analysis speed of the DNA microarray can be greatly improved, and even if the array lattice is inclined due to an error in the manufacturing process, the error of the result can be almost eliminated. Because it maintains the shape of, the error rarely occurs, thereby increasing the practical use of the DNA microarray's automatic analysis algorithm.

Hereinafter, exemplary embodiments of the present invention will be described with reference to the accompanying drawings.

2 is an operation flowchart of a method for analyzing a DNA microarray, and is composed of four steps as shown.

In step 1, canny edge detection is performed on a given image. As a result of the process, the boundary components of the probe spot of the image remain. This image has most of the boundaries for the actual probe spots, but it also contains all the boundaries of the noise present in the image and cannot be used as such.

In the second step, all image components with weak intensity are removed. In general, noise is often weaker than the actual image component. Therefore, the amount of noise disappears just by removing the weak components.

In the third step, all the removed and unclosed open figures are regarded as noise and deleted. Noise is often small in size, and very rarely has a closed figure that divides the inner and outer parts like the boundary of a probe spot. Therefore, after removing the component having the shape of an open figure, there is almost no remaining. As a result of the above step 3, an array image is obtained in which there are vacant places. (Figure 3)

In step 4, block seperation is performed on the array image having the empty space and divided into appropriate sizes. Hough transform is performed on each image obtained in the above process (FIG. 4) to obtain a linear component of the overall block, and then reconstruct the grid using the linear components. (FIG. 5). When the image width is N, the Hough transform

Since the process has a time function proportional to, divide it into K blocks through block seperation.

You can see a time gain proportional to. The probe spot is restored to the intersection point of the grid reconstructed by the Hough Transform, extraction of staining degree information from the spot is completed, and the analysis is completed.

1 is a reference diagram showing the manufacturing process of the DNA microarray

2 is a block diagram showing a method for analyzing DNA microarrays

3 is an image of an array of voids obtained in the midst of implementation of the invention.

4 is a result of the Hough Transform for FIG.

FIG. 5 shows a restored microarray using FIGS. 2 and 3.

Claims

Extracting edge components using canny edge detection from a DNA microarray image; A second step of recognizing the weak boundary component as noise and removing the removed image from the boundary component; A third step of obtaining an image from which the open figure is removed from the image obtained in step 2; Fully automated DNA microarray analysis method comprising the fourth step of recovering the normal spot lost in the image obtained in the third step to obtain the DNA probe staining.

The method of claim 1, wherein the analysis is performed without error on the tilted microarray image using the Hough transform in the step of restoring the normal spot in the fourth step.

The DNA microarray analysis method of claim 2, wherein, when the Hough transform is applied, a large image is segmented by block seperation of a large image into sections.