AssemblyQuality

Evaluate the quality of a de novo genome assembly using a variety of metrics. The input assembly may come from SGA or another assembler.

The quality metrics are partially based on Assemblathon 2: evaluating de novo methods of genome assembly in three vertebrate species. GigaScience 2013, 2:10. Metrics are computed separately for contigs and scaffolds. Some metrics (denoted R) are only computed when the reference genome is provided. These metrics are computed for the whole genome, and some (C) also individually for each reference chromosome. Metrics are as follows:

Basic contig/scaffold length statistics: {Total,Min,Max,Mean,Median}Length, NumSequences, LengthSD.
N50 and its generalization N(X) for X=1..100. All N(X) values are plotted and selected values are also in the tables.
Coverage (R): fraction (0..1) of reference genome that was assembled.
Validity (R,C): fraction (0..1) of assembly that was aligned to the reference.
Multiplicity (R): length of aligned assembly compared to length of covered reference regions; whether repeats were replicated or collapsed.
Parsimony (R,C): multiplicity/validity = length of scaffolds vs. length of covered reference regions.

Version	0.1
Bundle	sequencing
Categories	Assembly
Authors	Kristian Ovaska (kristian.ovaska@helsinki.fi)
Issue tracker	View/Report issues
Requires	Scala ; BWA
Source files	component.xml function.scala
Usage	Example with default values AssemblyQuality( contigs = FASTA, scaffolds = FASTA, //reference = FASTA )

Inputs

Name	Type	Mandatory	Description
contigs	FASTA	Mandatory	Assembled contigs.
scaffolds	FASTA	Mandatory	Assembled scaffolds.
reference	FASTA	Optional	Reference genome. If present, contigs and scaffolds are assembled to this genome and additional metrics are computed. The FASTA file must be accompanied with a BWT index created with "bwt index", in the same directory. This can be done with ReferenceIndexer component. For performance, a FASTA index (in Samtools format) is also recommended.

Name	Type	Description
report	Latex	Assembly coverage vs. contig/scaffold length plots (generalization of N50).
contigStats	CSV	Contig statistics that do not depend on the reference. There are two columns: Statistic (name of the metric) and Value.
scaffoldStats	CSV	Scaffold statistics that do not depend on the reference. There are two columns: Statistic (name of the metric) and Value.
contigReferenceStats	CSV	Contig statistics that depend on the reference (R): coverage, validity, multiplicity and parsimony. There are three columns: Statistic, Chromosome and Value. Statistics for all chromosomes combined are in rows with Chromosome=ALL. If the reference is not given, this file is empty.
scaffoldReferenceStats	CSV	Scaffold statistics that depend on the reference (R): coverage, validity, multiplicity and parsimony. There are three columns: Statistic, Chromosome and Value. Statistics for all chromosomes combined are in rows with Chromosome=ALL. If the reference is not given, this file is empty.

Test case	Parameters▼	IN contigs	IN scaffolds	IN reference	OUT report	OUT contigStats	OUT scaffoldStats	OUT contigReferenceStats	OUT scaffoldReferenceStats
case1	(missing)	contigs	scaffolds	reference	(missing)	(missing)	(missing)	(missing)	(missing)
case2_noref	(missing)	contigs	scaffolds	(missing)	(missing)	(missing)	(missing)	(missing)	(missing)
case3_empty	(missing)	contigs	scaffolds	reference	(missing)	contigStats	scaffoldStats	contigReferenceStats	scaffoldReferenceStats

Generated 2019-02-08 07:42:21 by Anduril 2.0.0