Genome Assembly of Amaranthus Hypochondriacus: An Emerging C4 Pseudo-Cereal Grain Crop

Amaranths are important pseudo-cereals native to the New World that have recently received increased attention due to their nutritional qualities and high seed protein content. We have sequenced the genome of Amaranthus hypochondriacus, a member of the Amaranthaceae family. We used the HiSeq Illumina platform to generate 180 bp paired-end reads for 150x coverage of the genome.  We also generated and sequenced 3Kb (135x coverage) and 6Kb (131x coverage) mate pair libraries for genome assembly scaffolding.  The ALLPATHS-LG assembler was used to assemble the reads into 23,420 contigs that were connected into 3,518 scaffolds with an N50 and L50 of 371 and 243Kb, respectively.  GapCloser V.1.12 (SOAP) was used to resolve N spacers introduced in the assembly scaffolding process reducing the total number of contigs from 23,420 to 17,366 which improved our contig N50 from 32,798 to 44,493 bp and reduced the total gap length by 7.3 Mb.  A 116x coverage BioNano physical map was generated and merged with the ALLPATHS-LG assembly to build 3,471 super-scaffolds, spanning 389.4Mb (~90%) of the estimated 432 Mb amaranth genome.  Thirty-six percent of the genome was estimated to be repetitive at K=25 scale.  The super-scaffold build had an N50 and L50 of 425Kb and 218, respectively, with the longest super-scaffold spanning 4.4 Mb.  Comparison of the super-scaffolds to the current SNP based linkage maps suggests that 1 cM is equal to ~123Kb in amaranth.