AltHap

By Abolfazl Hashemi, Banghua Zhu, and Haris Vikalo.

Table of Contents

0. Introduction
1. Citation
2. MATLAB implementation
3. Python implementation
4. Input file format
5. Output file format

Introduction

AltHap is a haplotyping assembly method for diploid, polyploid, and polyallelic polyploid organisms. The method formulates the haplotype assembly problem as a sparse tensor decomposition proplem and exploits the special structures of the factors. AltHap is implemented in both Python and MATLAB.

Note

The current implementations assume that the input file contains a connected component. The preprocessing step to handle disconnected componenets will be added in near future.

Citation

If you use AltHap in your research, please cite:

    @article{hashemi2017sparse,
        title={Sparse Tensor Decomposition For Haplotype Assembly Of Diploids And Polyploids},
        author={Hashemi, Abolfazl and Zhu, Banghua and Vikalo, Haris},
        journal={bioRxiv},
        pages={130930},
        year={2017},
        publisher={Cold Spring Harbor Labs Journals}
    }

MATLAB implementation

1. Usage in terminal:

matlab -nojvm -r "AltHap(ploidy,'input_file.txt','output_file.txt');exit"

The first two inputs are mandatory. If output_file is not given, the results will be saved in AltHap-output.txt. Please refer to the source code for a detailed explanation of other options which are mainly related to the stopping criteria. Please refer to the end of this file for a detailed explanation of input and output files format.

2. Run a simple example:

In this example, sim0.txt is the input fragment file, ploidy = 3, and the results are saved in AltHap-output.txt. There are 3 ways to run this example:

• In terminal write: matlab -nojvm -r "AltHap(3,'sim0.txt');exit"

• In terminal write: matlab -nojvm -nodesktop < ./simple_example.m

• Run the AltHapSimple.sh file in terminal. This file would then run the simple_example.m file.

Python implementation

Input file format

The input file format is as follows

Number of reads
Number of columns
Number_of_contiguos_segments Read_identifier Position_of_first_SNP_segment Continuous_bases_in_read Position_of_next_SNP_segment Continuous_bases_in_read ..... Quality_scores (in fastq format)

• Example for Biallelic:

5568
22801
2 chr22_SPA9_8733 2 0 5 0 ==
1 chr22_SPH2_1940 3 100 C==

• Example for Polyallelic:

4500
1000
2 chr3_1 1 13103 37 1132 IIIIIIIII
2 chr3_2 1 1310321 46 110302 IIIIIIIIIIIII
2 chr3_3 1 3220001 44 3001 IIIIIIIIIII
2 chr3_4 1 322000 40 30023 IIIIIIIIIII

Output file format

The output file contains MEC score, CPU Time, and Recovered Haplotypes. Subsequently, the phased haplotype is printed in the following format: first haplotype second haplotype third haplotype ...

• Example for Biallelic:

MEC: 353
CPU Time: 38.363051
Recovered Haplotype:
0 1 1
1 0 1
0 0 1
1 1 1
0 1 1
1 0 0
1 0 1
1 1 1

• Example for Polyallelic:

MEC: 51
CPU Time: 28.912201
Recovered Haplotype:
1 1 1
1 0 0
3 3 1
2 2 0
1 3 1
3 0 0
0 0 2
3 2 0

For higher ploidy, there are K phased haplotypes instead of 3.