Recombination is the process where two DNA molecules exchange nucleotide sequences with each other. The existence of recombination hotspots offers a way to learn what other processes are associated with recombination. The objective of our work is to find a better predicting model for recombination hotspot. iRSpot starts with DNA sequences for given hotspot and coldspot dataset. We use three feature extraction technique to find important features. Recursive feature elemination and XGboost both are used for feature selection. Model gives 77% accuracy after applying 1D neural network.
This project is supervised by Md Rakibul Haque Sir (faculty member, Department of CSE, United International University)