3.4 | Genome assemblies of mountain nyala (degenerated DNA)
Mate-pair generation of T. buxtoni , using B. grunniens as a reference, yielded the maximum number of mate pairs (B. grunniens : 416,044,705) while using M. moschiferus produced the least number of mate pairs (M. moschiferus : 220,576,118). The number of mate pairs generated using B. grunniens (same subfamily) as the reference genome was greater than that using T. scriptus and T. strepsiceros (same genus) as reference genomes (T. scriptus : 305,670,717, T. strepsiceros : 392,062,745), and this may be attributed to the high quality of B. grunniensassembly (Table S7-S9).
The mountain nyala (T. buxtoni ) genome, which was generated with only paired-end reads from the degenerate samples, was not well assembled (Chen et al., 2019). The quality of the draft genome generated without using in silico mate-pair libraries was unsatisfactory (N50: 3.5 Kb, complete BUSCOs: 645) (Table 3). Therefore, we used the original as well as the optimized in silico method to perform genome assembly of the mountain nyala. The results showed that when the original mate pairs were generated using different references (‘scr’:Tragelaphus scriptus , ‘str’: Tragelaphus strepsiceros , ‘gru’: Bos grunniens , ‘mos’: Moschus moschiferus ), the draft genomes were improved, showing higher contiguity (N50–scr*: 592 Kb, str*:431 Kb, gru*:2.6 M, mos*:1.5 M) and increased completeness (Complete BUSCOs: scr*:1,956, str*:1,979, gru*:2,018, mos*:1,697). Compared to assemblies using the in silico method, genomes assembled using conserved mate pairs did not increase N50 (scr-str**: 203 Kb, gru-scr**: 474 Kb) or the number of complete BUSCOs (scr-str**: 1,727, gru-scr**: 1,759). Due to the low quality of the mountain nayala genome, no good reference genome could be used to calculate the misassembly rate.