PTG: Pasimonious tree-grow method for haplotype inference

• Last updated: May 22, 2005

Reference#

• Zhenping Li, Wenfeng Zhou, Xiangsun Zhang and Luonan Chen. A parsimonious tree-grow method for haplotype inference. Bioinformatics, Vol. 21, No. 17, pp. 3475-3481, 2005.

Method#

PTG (Pasimonious Tree-Grow method) program has four options to infer haplotypes from genotype data based on parsimony principle.

Simulation on model 1#

1. Input the number_of_genotypes m, which is the rows of genotype matrix.
2. Input the number_of_SNPs n, which is the number of SNP sites.
3. Click inputH, input 2m haplotypes in the Input_data window, and click OK. Then the genotype matrix in the main windows is generated by conflate the 2m haplotypes of the input data, where the (2i-1)-th haplotype and the (2i)-th haplotype conflate the i-th genotype.
4. Click button "Run", the result will be given in the "haplotype result" and "index set" windows. The total CPU time will be shown in the windows.
5. Click button "error". The error rate will be calculated in the corresponding windows, where the error_rate on the left top is the rate of genotypes that is not correctly inferred, i.e., the haplotypes inferred by PTG are not the same as those of input data.
6. Click the button of "save_result", the result will be saved in a txt document.

Simulation on model 2#

1. Input the number_of_genotypes m, which is the rows of genotype matrix.
2. Input the number_of_SNPs n, which is the number of SNP sites.
3. Click "inputH"
4. Input the number of the dictinct haplotypes in the "Haplotype no" window.
5. Input the distinct haplotypes and the corresponding frequency in the "Haplotype" and the "frequency" windows respectively and then click "Haplotype" and OK.
6. Click the button of "Haplotype" to generate the 2m haplotypes and the corresponding genotype matrix. The genotype matrix will be showed in the main window. Then click the OK button to close the subwindow.
7. Click "Run" button to execute PTG program.
8. Click "error" Button to see the results. The input haplotypes will be shown in the windows of "H0", and the inferred haplotypes by PTG will be shown in the windows of "H1". On the top of the windows, BN denotes the number of distinct haplotypes for input data, whereas EN denotes the number of distinct haplotypes inferred by PTG. Click OK to close the "error" window.
9. Click the button of "save_result", the result will be saved in a txt document.

Simulation on model 3#

1. Input the number_of_genotypes m, which is the rows of genotype matrix.
2. Input the number_of_SNPs n, which is the number of SNP sites.
3. Click "auto_generation": the genotype matrix will be generated by randomly generating 2m haplotypes and conflating them.
4. Click button "Run", the result will be given in the "haplotype result" and "index set" windows. The total CPU time will be shown in the windows.
5. Click button "error". The error rate will be calculated in the corresponding windows, where the error_rate on the left top is the rate of genotypes that is not correctly inferred, i.e., the haplotypes inferred by PTG are not the same as those of input data.
6. Click the button of "save_result", the result will be saved in a txt document.

Notice that when clicking "auto_genenration": the genotype matrix will be provided by randomly generating 2m haplotypes and conflating them. However, such data may have no similarity with real haplotypes. The error rate may also be high, when the recombination rate is very high in the randomly simulation data.

Download#

• PTG.exe
• Supplementary Information