A binning tool to reconstruct viral haplotypes from assembled contigs

Jiao Chen; Jiayu Shang; Jianrong Wang; Yanni Sun

First, we constructed 5 haplotypes with high sequence similarity. Second, in order to simulate haplotypes of different relative abundances, we generated 3 sets of reads following different abundance distributions. Third, we generated contigs using two different methods. In method 1, we simulated 5 sets of error-free contigs of different sizes directly from the reference genomes. In method 2, we applied available assembly tools to generate contigs from the reads. The simulated contigs are not dependent on any assembly tool and thus are ideal for evaluating the binning method. For simulated contigs, we have 3 (sets of reads) × 5 (sets of simulated contigs), i.e., 15 sets of input to our program. For assembled contigs, we have 3 (sets of reads) × 2 (sets of assembled contigs) as input to VirBin because we applied two assembly tools. Figure S1 in the Supplementary Data file sketches the process of input data generation. The data simulation details can be found below.