Research On The Triploid Individual Haplotype Reconstruction Problem

Studying SNP plays an important role in exploring genetic traits and phenotypic differences of triploid species,but haplotype contains more genetic information than a single SNP site,which plays a more important role in studying phenotypic differences,diseases prediction and gene expression.Due to the limit of current experimental technology,inferring haplotypes is both time consuming and expensive by using biological experimental directly,thus,obtaining individual haplotypes by computational methods become the hot topic in bioinformatics.This paper mainly studies the triploid individual haplotype reconstruction problem and the specific work is as follows:For the minimum error correction with genotype information model put forward a reconstruction algorithm HTMS(Haplotyping a Triploid individual with Most Support)based on support degree.The HTMS algorithm reconstructs the SNP sites of the three haplotypes one after another.When reconstructing a given SNP site,it enumerates three kinds of SNP values in terms of the genotype of the site,and chooses the one with the most support coming from the SNP fragments that are covering the corresponding SNP site.In the experiments,two kinds of simulators CELSIM and MetaSim were invoked to generate SNP fragments.The reconstruction rate and running time were compared and analyzed among algorithms HTMS,T-HC,GTIHR,W-GA and Q-PSO with different parameters settings,such as fragment coverage,error rate,single fragment length,haplotype length and haplotype hamming distance.Under different parameter settings,the HTMS algorithm can obtain highest reconstruction rate under fastest running speed,which were proved by a number of experiments.For the minimum error correction with genotype information model,a reconstruction algorith:m HTLD(Haplotyping a Triploid individual with Least Difference)based on difference degree is proposed.HTLD algorithm uses a new method to measure the distance between fragments and haplotypes.When enumerate the three values of the SNP site under reconstruction based on the genotype,the values are calculated respectively by haplotype and fragments which covers the site,and select the corresponding to the least difference degree of value,which corresponds to the smallest distance value.Algorithm HTMS and HTLD have similar performance,and comparing with algorithms T-HC、GTIHR、W-GA and Q-PSO,they can obtain higher precision with the shorter running time under different parameter settings,which was tested by a number of experiments.To sum up,based on the minimum error correction with genotype information model,this paper proposes two reconstruction algorithms HTMS and HTLD.The experiments results show that the two algorithms can reconstruct the haplotypes that have higher precision with faster running speed.They are effective methods to solve triploid individual haplotype reconstruction problem and are very practical for realistic applications.