The genome of the tegu lizard Salvator merianae: combining Illumina, PacBio, and optical mapping data to generate a highly contiguous assembly

Abstract

Background: Reptiles are a species-rich group with great phenotypic and life history diversity but are highly underrepresented among the vertebrate species with sequenced genomes.

Results: Here, we report a high-quality genome assembly of the tegu lizard, Salvator merianae, the first lacertoid with a sequenced genome. We combined 74X Illumina short-read, 29.8X Pacific Biosciences long-read, and optical mapping data to generate a high-quality assembly with a scaffold N50 value of 55.4 Mb. The contig N50 value of this assembly is 521 Kb, making it the most contiguous reptile assembly so far. We show that the tegu assembly has the highest completeness of coding genes and conserved non-exonic elements (CNEs) compared to other reptiles. Furthermore, the tegu assembly has the highest number of evolutionarily conserved CNE pairs, corroborating a high assembly contiguity in intergenic regions. As in other reptiles, long interspersed nuclear elements comprise the most abundant transposon class. We used transcriptomic data, homology- and de novo gene predictions to annotate 22,413 coding genes, of which 16,995 (76%) likely have human orthologs as inferred by CESAR-derived gene mappings. Finally, we generated a multiple genome alignment comprising 10 squamates and 7 other amniote species and identified conserved regions that are under evolutionary constraint. CNEs cover 38 Mb (1.8%) of the tegu genome, with 3.3 Mb in these elements being squamate specific. In contrast to placental mammal-specific CNEs, very few of these squamate-specific CNEs (<20 Kb) overlap transposons, highlighting a difference in how lineage-specific CNEs originated in these two clades.

Conclusions: The tegu lizard genome together with the multiple genome alignment and comprehensive conserved element datasets provide a valuable resource for comparative genomic studies of reptiles and other amniotes.

Figure 1:

Workflow to generate the tegu lizard v2 assembly. (A) The tegu lizard, Salvator merianae. (B) Assembly v1 was built entirely from Illumina short-read data. To improve this assembly, we used Pacific Biosciences (PacBio) long-read data to close assembly gaps and extend scaffolds and merged the improved Illumina with a PacBio-only assembly. Finally, optical mapping data were used to resolve contig chimeras and scaffold even further. Used tools and their input data are shown in gray.

Figure 2:

Comparison of assembly contiguity. N(x)% graphs show the contig (A) and scaffold (B) sizes (y-axis), where x% of the genome assembly consists of contigs and scaffolds of at least that size. The tegu lizard v1 and v2 assemblies are shown in gray and black. All other assemblies are sorted by the N50 values in the insets. Dashed lines mark the N50 and N90 values.

Figure 3:

Comparison of genome completeness for coding genes. The bar charts show the number of complete, fragmented, and missing genes using two BUSCO datasets for vertebrate-conserved (left) and tetrapod-conserved (right) genes.

Figure 4:

Using conserved non-coding elements to compare genome completeness and contiguity. Bar charts show (A) the number of aligning UCEs that do not overlap coding regions (N = 197 in total) and (B) the number of aligning CNEs that do not overlap exons (N = 493 in total). (C) The percentage of 282 evolutionarily conserved pairs of neighboring CNEs that are also found as neighbors in the squamate assemblies. Both UCE and CNE sets are highly conserved among vertebrates and thus are likely to exist in squamates.

Figure 5:

Repeat landscape in squamate genomes. Major classes of repeats are color coded and shown as bar charts that represent the portion of the genome they cover. Simple repeats comprise tandem repeats, low complexity regions, and satellite repeats.

Figure 6:

Phylogenetic tree of the amniote species included in our multiple genome alignment. The topology of the tree is based on [46] and [47]. Branch lengths represent the number of substitutions per neutral site, as estimated from four-fold degenerated codon positions.

Figure 7:

Transposon-derived conserved non-exonic (CNE) squamate-specific elements. (A) A squamate-specific CNE likely originated from the insertion of a short interspersed nuclear element (SINE) belonging to the Mammalian-wide interspersed repeats (MIR) family. The multiple genome alignment shows that this CNE is highly conserved among squamates but does not align to non-reptile species. (B) Several squamate-specific CNEs likely originated from the insertion of a LINE of the RTE-BovB family. This insertion likely happened after the split from the lineage leading to geckos as no sequence aligns to the gecko genome.