A comparative evaluation of genome assembly reconciliation tools

Table 1 Features of the assembly reconciliation tools evaluated in this study

	CISA	GAA	GAM_NGS	GARM	Metassembler	MIX	ZORRO
Inputs
Contigs allowed	✓	✓	✓^a	✓	✓	✓	✓
Scaffolds allowed	✓	✓^b	✓^a	✓	✓
Short reads allowed			✓^a
Paired-end reads allowed			✓^a				✓
Mate-pair reads allowed			✓^a		✓
Alignments allowed		✓	✓
Reads required			✓^a		✓		✓
Reference input assembly required			✓		✓
Input assemblies treated symmetrically	✓					✓
Only two input assemblies		✓	✓	✓	✓^c		✓
More than two input assemblies	✓					✓
Can handle bacterial/small genomes	✓	✓	✓	✓	✓	✓	✓
Can handle large eukaryotic genomes		✓	✓	✓	✓		✓
Goals
To increase assembly contiguity	✓	✓	✓	✓	✓	✓	✓
To decrease number of assembly errors	✓		✓
Methods
Compression-expansion statistic		✓	✓		✓
Scaffolding information		✓		✓
Use single reads			✓
Use paired-end/mate-pair reads		✓	✓		✓		✓
Can split assembly misjoin	✓	✓
Can detect/avoid repetitive regions	✓		✓				✓
Output
Contigs				✓	✓	✓
Scaffolds	✓	✓	✓	✓^d	✓	✓	✓

^aOptional, as GAM_NGS requires an alignment file
^bScaffolds should be broken into contigs. A gap file and contig naming convey scaffolding information
^cPerforms iterative pairwise
^dWhen the input contains scaffolds

ISSN: 1474-760X