Meiotic recombination caused by meiotic double-strand DNA breaks. In some areas the frequency of gene recombination is relatively high, while in other regions the frequency is lower, we called them hotspots and coldspots, respective. Hot/Cold spots have given clues as to the mechanism of recombination initiation, so, it is important to accurate prediction of hot/cold spots.
In this study, we rebuild a equilong dataset base on posted data sets, then we extract the eigenvector with the 6-mers, binomial distribution as a feature selection technology was proposed in this work. Then the SVM classifier was employed. The 5-fold cross-validated results showed that the maximum overall accuracy of 84.08% was achieved when the optimal 381 6-mers was used.
