The process of mapping large numbers of markers is computationally complex, as the increase of numbers of markers results in an exponential increase in the mapping runtime. Also, having unreliable markers in the dataset adds more complexity to the mapping process. In this research, we have addressed these two issues and proposed our solution. The proposed approach builds solid maps in two phases: Phase 1 builds an initial map following these steps: 1) Resample the original dataset to generate variant datasets, then cluster all resampled datasets into groups of markers. 2) Merge all groups of markers to filter out unreliable markers. 3) Generate a Map for each group of markers. 4) Concatenate all groups' maps to form the final map. Phase 2, Adds more markers to the initial framework to build a high resolution map as follows: 1) Use Kmeans algorithm to filter out unreliable markers and cluster the remaining markers. 2) Insert the remaining markers in their best positions in the initial framework. To evaluate the performance of the proposed approach, we compare our constructed maps on the human genome with the physical maps. Moreover, we compare our constructed maps with a state-of-the-art tool for building maps. Experiment results show that the proposed approach has a very low computational complexity and produces solid maps with high agreement with the physical maps.

No abstract

Abstract-The process of mapping large numbers of markers is computationally complex, as the increase of numbers of markers results in an exponential