Accession Number:

AD0808715

Title:

MATCH RECOGNITION IN CHROMOSOME BAND STRUCTURE.

Descriptive Note:

Technical rept.,

Corporate Author:

TEXAS UNIV AT AUSTIN LABS FOR ELECTRONICS AND RELATED SCIENCE RESEARCH

Personal Author(s):

Report Date:

1967-01-15

Pagination or Media Count:

65.0

Abstract:

For over thirty years research geneticists have been studying the chromosomes of the species Drosophila fruitfly. These chromosomes, suitably prepared and stained, have an appearance similar to earthworms with light and dark bands around them. They are large enough to be seen clearly under a high power light microscope. These patterns of bands reflect the genetic heritage of the fly and can be used to trace inherited traits. Researchers examine enlarged photomicrographs of the chromosomes to locate sections with closely similar band patterns. An application of pattern recognition techniques to this problem has yielded a semi-automatic method of detecting band patter similarities. The chromosome photographs are optically scanned in a specially constructed densitometer. This generates a plot of optical density versus position along the length of the chromosome. The density plot is measured to determine the widths and spacings of the bands. A sequence of band widths and spacings is used as the mathematical representation of the chromosome. These sequences are used as input data for a computer program which locates matching sections. The computer program uses a method which was designed by inspection of known band matches. It checks the similarity of all combinations of sections and prints out the location of those which match. The result is a listing of probable matching sections. Preliminary testing on a number of chromosomes indicates that the match recognition accuracy is as high as 90 on data from good quality photographs. It is expected that this method will be of considerable aid to researchers interested in Drosophila chromosome band structure. Author

Subject Categories:

  • Biology
  • Anatomy and Physiology
  • Computer Programming and Software

Distribution Statement:

APPROVED FOR PUBLIC RELEASE