This paper describes some parallel processing algorithms for the analysis of two-dimensional electrophoresis images. The machine used for the processing was the CLIP4 Cellular Array Computer at University College, London, one of the largest processor arrays in the world. Included in this paper are an algorithm for centroid detection, Gaussian fitting algorithms, and an algorithm for the extraction of data out of the cellular array machine. It is shown that these parallel algorithms can run at a speed almost completely independent of the number of spots in the gel images. 8 1986 Academic press, hc.
Advances are being made in the field of computer-assisted analysis of 2D electrophoresis images. As well as new improved algorithms for the analysis of data, especially in gel matching algorithms, powerful image processing computers have been introduced to perform the analysis at much greater speeds than obtainable by conventional machines. These machines use parallel algorithms to obtain the improved efficiency. Skolnick (1) used the pipeline processing machine CYTOCOMPUTER to detect the spot peaks in the gel image. The author, in a previous article in this journal (2), described the main analysis algorithms using the CLIP4 cellular array computer, one of the largest processor arrays in the world. It is only useful to use these type of machines if efficient algorithms are developed which utilize them to their full potential.
In this paper, some more parallel algorithms for the analysis of the gels are described. The reader is referred to the referenced article (2) for a description of the CLIP image processing system and basic analysis algorithms. However, a brief summary of that paper is presented below.
THE CLIP SYSTEM AND BASIC ANALYSIS ALGORITHMS
CLIP4 is a parallel cellular array computer made up of an array of 9216 processors in a % x 96 grid. Each processor can communicate with its eight neighbors.