Haplotype-resolved genome of heterozygous African cassava cultivar TMEB117 (Manihot esculenta)

Abstract

Cassava (Manihot esculenta Crantz) is a vital tropical root crop providing essential dietary energy to over 800 million people in tropical and subtropical regions. As a climate-resilient crop, its significance grows as the human population expands. However, yield improvement faces challenges from biotic and abiotic stress and limited breeding. Advanced sequencing and assembly techniques enabled the generation of a highly accurate, nearly complete, haplotype-resolved genome of the African cassava cultivar TMEB117. It is the most accurate cassava genome sequence to date with a base-level accuracy of QV > 64, N50 > 35 Mbp, and 98.9% BUSCO completeness. Over 60% of the genome comprises repetitive elements. We predicted over 45,000 gene models for both haplotypes. This achievement offers valuable insights into the heterozygosity genome organization of the cassava genome, with improved accuracy, completeness, and phased genomes. Due to its high susceptibility to African Cassava Mosaic Virus (ACMV) infections compared to other cassava varieties, TMEB117 provides an ideal reference for studying virus resistance mechanisms, including epigenetic variations and smallRNA expressions.

Fig. 1

Overview of the cassava cultivar TMEB117 genome. (a) Circos plot displays repeat and gene densities for the two haplotypes visualized in 1 Mbp sliding windows. The tracks from the outer to inner show, (i) Repeat density for hap1 genome (ii) Gene density for hap1 genome (iii) Repeat density for hap2 genome (iv) Gene density for hap2 genome. (b) Cassava plant in a pot from the screen house. (c) BUSCO score of the TMEB117 genome.

Fig. 2

Illustrate the proportion and distribution of TEs across the chromosomes, as annotated by EDTA. (a) Shows the proportion of TEs identified in the hap1 genome, with the most abundant type LTR-RT (represented in the blue segment in the pie chart), covering 57.37% of the genome. (b) In hap2, LTR-RT remains the predominant TE family covering 55.07% of the genome. (c) Provides an overview of the distribution of TE families across all cassava chromosomes. (d) Slight difference in the distribution of TEs families annotated in the chromosomes of the hap2 genome compared to the hap1 genome.

Fig. 3

Venn diagram of the number of gene families shared among and unique to the haplotype genomes of three African cassava cultivars: TMEB117 (hap1 and hap2), TME204 (hap1 and hap2), TME7 (hap1 and hap2), in comparison to the reference genome AM560-2 v8. (a) 18,770 core gene families shared among the first haplotigs comparison with the reference genome AM560-2 v8. The second comparison (b)19,588 core genes on the second haplotig comparison with the reference genome AM560-2 v8. 931 gene families were unique in TMEB117 hap1 genome and 1042 in the hap2 genome.

Fig. 4

The completeness of resolved haplotypes assessed by Merqury copy number spectrum plots (a) and assembly plots (b). The x-axis represents the k-mer multiplicity, while the y-axis shows the abundance of k-mers. The grey region. represents the abundance of k-mers in the HiFi reads missing in the scaffold of the genome. (a) Copy number spectrum plot - the red peak observed at ~ 25x indicates heterozygous k-mers (1-copy k-mers), while the blue peak at ~ 50x represents the homozygous k-mers (2-copy k-mers). The other peaks show low levels of duplicated k-mers. (b) Assembly plot – k-mers coloured by their uniqueness, red peak (hap1), blue peak (hap2). At the heterozygous peak (25x), there is a slight difference in the k-mers indicating reconstruction of heterozygous variants was almost complete. Shared k-mers are shown in green which is at the 50x k-mer multiplicity.