Apis mellifera graph genome

AmelGraph 1.1.0

We aligned 5 different publicly available assemblies with the cactus pangenome workflow (v2.0.5):

Use 

Species 

Genome ID 

Accession 

Graph ID 

Size (Mb) 

Reference 

A. mellifera (DH4) 

Amel_HAv3.1 

GCF_003254395.2 

DH4 

225.2 

Derivate 

A. m. mellifera 

INRA_AMelMel_1.0 

GCA_003314205.1 

mellifera 

227.0 

Derivate 

A. m. carnica 

ASM1384124v2 

GCA_013841245.2 

carnica 

226.0 

Derivate 

A. m. caucasica 

ASM1384120v1 

GCA_013841205.1 

caucasica 

224.8 

Derivate 

A. m. ligustica 

ASM1932182v1 

GCA_019321825.1 

ligustica 

231.1 

We used the cactus1 pangenome workflow to generate a 5-ways pangenome alignment, masking in blocks of 10Kb, and considering all the sequences in the different data. Due to incompatibilities with the pangenome workflow generation of the indexes, we regenerated the giraffe indexes from the output GFA/VCF file. We also downloaded annotations available for three genomes (DH4, A. m. carnica and A. m. caucasica), modifying the naming of the contigs to include the subspecies name (i.e. 'LG1' for A. m. caucasica modified to 'caucasica.LG1'). We then ran the vg autoindex function:

vg autoindex -w giraffe -o pangenome -t 8 -T ./TMP -x carnica.gff -x caucasica.gff -R XG -x DH4.gff

We compared this 'full' cactus graph to one built using only sequence on the linkage groups, and to one generated using the PGGB workflow. We evaluated the different graphs using PGGE. This full cactus graph returned the highst aligned identity, sequence matches, and unique alignments, while having the lowest number of multiple mapping and missing alignments.

Parameter 

CACTUS (LG) 

CACTUS (FULL) 

PGGB (norm) 

HAv3.1 size 

221,626,419 

225,250,884 

225,250,884 

Graph length (bp) 

243,077,200 

246,675,144 

315,672,860 

Extra sequence 

21,450,781 

21,424,260 

90,421,976 

# nodes 

17,120,383 

11,664,080 

17,952,597 

# edges 

21,385,634 

15,990,593 

21,702,958 

References

1.  Armstrong, J. et al. Progressive Cactus is a multiple-genome aligner for the thousand-genome era. Nature 587, (2020).