sample. The pile-up is then used to identify regions where the
mapped reads disagree with the reference genome in a process
called variant-calling, typically generated with tools like samtools
[59] or the GATK package [60]. A list of variants between the
sample read set and the reference genome is summarized in a textbased VCF file. A variant for a given DNA base can be one or more
alternative bases (i.e., a SNP) or an indel (e.g., a gap). Finally, if a
consensus genome sequence is desired based on these called variants, one can designate a set of rules for generating such a consensus (such as calling ambiguous bases for SNPs with >1 alternative
base) or coverage required to report a base (typically at least 10Â,
or a base will be reported as “N” or “?”). This can be done using the
bcftools [59] consensus tool.
4 Species-Level Identification
To understand how WGS can be used for species-level identification of a sequenced genome, it is first important to understand
perhaps the most basic question of all in bacterial taxonomy: What
is a species? And, the obvious question that follows from this one:
How do we identify species and assign strains to species?
Each species is represented by a type strain and a description of
that strain. Usually, the type strain is the first strain identified;
however, it is not necessarily the most typical or representative of
the species. If the original strain upon which the description was
based cannot be found or is no longer appropriate, a neotype strain
may be proposed. As a pertinent example, Mycobacterium tuberculosis was identified in 1882 by Robert Koch. Strain H37Rv was
proposed as the neotype strain in 1972 [61] and later accepted.
Now, Mycobacterium tuberculosis H37Rv
T is the type strain of the
species. As the only strain with official standing in nomenclature,
the type strain is in essence the “definition” of a species.
A review of bacterial and archaeal species definitions recommended a species concept that should allow for “unequivocal identification” into categories that circumscribe “monophyletic, and
genomically and phenotypically coherent populations of individuals
that can be clearly discriminated from other such entities by means
of standardized parameters” [16]. The parameters generally used
constitute what is termed the polyphasic approach.
In order to identify a strain and assign it to a species, the
characteristics of the strain in question are compared to the characteristics of various species type strains. These type strains serve as
the taxonomic reference points for comparative studies. A strain
that shares enough of the essential characteristics of a type strain is
said to be within the circumscription of that type strain. Therefore,
it belongs to that species which is represented by that type strain.
Naturally, there is some flexibility in characteristics in order to
418
Marco A. Riojas et al.
mapped reads disagree with the reference genome in a process
called variant-calling, typically generated with tools like samtools
[59] or the GATK package [60]. A list of variants between the
sample read set and the reference genome is summarized in a textbased VCF file. A variant for a given DNA base can be one or more
alternative bases (i.e., a SNP) or an indel (e.g., a gap). Finally, if a
consensus genome sequence is desired based on these called variants, one can designate a set of rules for generating such a consensus (such as calling ambiguous bases for SNPs with >1 alternative
base) or coverage required to report a base (typically at least 10Â,
or a base will be reported as “N” or “?”). This can be done using the
bcftools [59] consensus tool.
4 Species-Level Identification
To understand how WGS can be used for species-level identification of a sequenced genome, it is first important to understand
perhaps the most basic question of all in bacterial taxonomy: What
is a species? And, the obvious question that follows from this one:
How do we identify species and assign strains to species?
Each species is represented by a type strain and a description of
that strain. Usually, the type strain is the first strain identified;
however, it is not necessarily the most typical or representative of
the species. If the original strain upon which the description was
based cannot be found or is no longer appropriate, a neotype strain
may be proposed. As a pertinent example, Mycobacterium tuberculosis was identified in 1882 by Robert Koch. Strain H37Rv was
proposed as the neotype strain in 1972 [61] and later accepted.
Now, Mycobacterium tuberculosis H37Rv
T is the type strain of the
species. As the only strain with official standing in nomenclature,
the type strain is in essence the “definition” of a species.
A review of bacterial and archaeal species definitions recommended a species concept that should allow for “unequivocal identification” into categories that circumscribe “monophyletic, and
genomically and phenotypically coherent populations of individuals
that can be clearly discriminated from other such entities by means
of standardized parameters” [16]. The parameters generally used
constitute what is termed the polyphasic approach.
In order to identify a strain and assign it to a species, the
characteristics of the strain in question are compared to the characteristics of various species type strains. These type strains serve as
the taxonomic reference points for comparative studies. A strain
that shares enough of the essential characteristics of a type strain is
said to be within the circumscription of that type strain. Therefore,
it belongs to that species which is represented by that type strain.
Naturally, there is some flexibility in characteristics in order to
418
Marco A. Riojas et al.