Research On Automatic Identification And Location DNA Fingerprinting Gel

The fundamental techniques involved in DNA fingerprinting were discovered serendipitously in 1984 by geneticist Alec J. Jeffreys of the University of Leicester in Great Britain while he was studying the gene for myoglobin, a protein that stores oxygen in muscle cells. The term"DNA fingerprinting"was coined to allude to the traditional use of fingerprints as the most unique mean of human identification. Now, DNA fingerprinting has become an indelible part of society, helping to prove innocence or guilt in criminal cases, resolving identifying victim and clarifying paternity. Furthermore, it has been used in the biology research widely, such as molecular biology research, the identification of the breed resource, crop heredity breeding, et al. But it is a boring job for biologists to analysis the DNA fingerprint, which will cost them an abundance of time and energy. However, because of the limit of analysis by human, the results which analyze by different person may be different. Due to the development of computer science, it is become possible to establish an automatic DNA fingerprint analysis system which can recognize image, analyze data automatically.The paper analyzes the structure and algorithm of the DNA fingerprinting automatic analysis system, establishes a mathematic model to location the restriction fragments, develops confidence interval method, which is based on threshold, to divide image.The paper presents a mathematical model for location and shape of the restriction fragments as a function of fragment size. Automated identification of restriction fragment involves several steps, include: image preprocessing, to get rid of noise and enhance image, for instance, histogram equalization, Laplace filter, Median filter, et al.; lane division, for division of the each lane of the fingerprint image. The purpose of this step is to assign allele band to a separate lane on the image. The column of pixels which has the most total intensity represents the optimal division between two allele lanes. A dividing line is drawn corresponding to that pixel column; lane analysis, for the last step of digital image processing. The task of the stage is to separate the every allele band from the background. The paper develops confidence interval method to separate the image correctly. The fundamental of the method is: the all top intensity of y orientation is regarded as a population which has a mean and a standard deviation. A confidence limit will be selected to determine the criterion, which is indeed the threshold. Comparing with the Otsu Method, confidence interval method avoids the"agglutinate"phenomenon efficiently.