Abstract:

The meiotic recombination events do not occur with a uniform frequency throughout the genome but with a higher rate in some regions and lower in others. Characterization and identification of meiotic recombination spots is critical for our understanding of the recombination mechanism. In this paper, we first propose a new 3-D graphical representation for a DNA sequences. Then, combining the 3-D graphical representaion with Z-curve, we characterize a DNA sequence by a 13-D vector whose components are the corresponding normalized ALE indices. Support vector machine (SVM) and Jackknife cross-validation test are employed to perform our method on a benchmark dataset for recombination spots. Results show that our method achieved an overall accuracy of 93.70% with the Matthew′s correlation coefficient (MCC) of 0.873, which suggests that the proposed method can serve as a useful tool for identifying the recombination spots.

Key words: 3-D graphical representation, ALE-index, meiotic recombination, support vector machine