Fig. 1

Schematic of gcaPDA workflow. (a) Gamete cells (N = 40) are isolated from focal individual. (b) Whole genome shotgun sequencing is performed to generate gamete cell short reads. (c) HiFi reads and Hi-C reads are generated by sequencing somatic tissues. (d) An initial assembly is generated by assembling HiFi reads into contigs and scaffolding contigs into superscaffolds with Hi-C data. (e) Short reads of gamete cells are then mapped to the initial assembly and (f) SNPs were identified for each gamete cells. (g) Chromosomal-scale haplotypes of the sequenced individual are reconstructed based on gamete cell SNP arrays using major voting strategy [23], with number of gamete cells that supports adjacent SNP combination were shown on the left. By comparing SNPs of each gamete cell with reconstructed haplotypes, (h) crossovers and (i) haplotype blocks of gamete cells can be determined. Gamete cell reads are (j) partitioned based on haplotype blocks and (k) normalized by k-mer depth to mimic genome coverage distribution of regular parental WGS reads. At last, (l) HiFi reads and partitioned normalized gamete cell reads are then used to construct phased contigs using hifiasm [11] and scaffolded into phased pseudochromosomes with Hi-C data