genome from the (relatively short) sequence
reads. Next, the genome assembly is annotated.
This entails the identification of repetitive
sequences, genes, and other functional elements in the genome. The predicted genes are
subsequently annotated by assigning a putative
function, usually based on homology to known
genes and domains.
This chapter will describe advances in technologies underlying fungal genome sequencing,
annotation, and analysis. Furthermore, the
impact of fungal genome sequencing is illustrated using examples from several fields of
biotechnology.
II. Advances in Genome Sequencing
Technologies
In the past decade, sequencing technologies
have improved dramatically, radically changing
the landscape of fungal genome sequencing.
Sanger sequencing was the first sequencing
technique that was used for genome sequencing
(Sanger et al. 1977b). It was used to sequence
landmark genomes such as the first bacteriophage jX174 (Sanger et al. 1977a), the first
bacterium Haemophilus influenza (Fleischmann et al. 1995), the first eukaryote (and first
fungus) Saccharomyces cerevisiae (Goffeau
et al. 1996), the first plant Arabidopsis thaliana
(Arabidopsis Genome Initiative 2000), the first
animal Caenorhabditis elegans (The C. elegans
Sequencing Consortium 1998), as well as the
human genome (International Human Genome
Sequencing Consortium 2001). The sequencing
of these genomes was generally a multi-year
undertaking and was performed by large consortia of collaborating labs. Sequencing reads
that were obtained with Sanger technology were
relatively long (up to approximately 1500 bp)
and were relatively straightforward to assemble
using assembly software such as Jazz (Aparicio
et al. 2002) or Arachne (Batzoglou et al. 2002).
More recently, since the mid-2000s, several
new sequencing platforms were developed that
are collectively known as “next-generation
sequencing” (NGS). Initially, these techniques
included the now mostly defunct technologies
Roche 454 (Margulies et al. 2005), IonTorrent
(Life Technologies) and SOLiD (Applied Biosystems). Currently the most prominent short
read sequencing technology, however, is Illumina (Bennett 2004). Although the sequences
generated by Illumina technology were initially
too short for efficient genome sequencing (up
to 25 bp), this has increased to currently
2 Â 300 bp on an Illumina MiSeq machine.
New assembly approaches and software were
developed for these short reads, such as Velvet
(Zerbino and Birney 2008), ABySS (Simpson
et al. 2009), SOAPdenovo (Luo et al. 2012),
and SPAdes (Bankevich et al. 2012).
Rather paradoxically, the assemblies generated from early NGS techniques were not nearly
as good as the ones generated from Sanger
reads, with respect to assembly fragmentation.
Especially repetitive genomic regions (e.g., originating from transposable elements) were
challenging to assembly using short reads.
However, crucial advantages of NGS technologies are that they are considerably faster and
cheaper than Sanger sequencing (Ghurye and
Pop 2019). This meant that genome sequencing
became affordable to core facilities and even
individual researchers, as opposed to the large
sequencing consortia that were required for
Sanger-based genome sequencing. This is illustrated by the following back-of-the-envelope
calculation: sequencing a typical fungal genome
of 30 Mbp with 100-fold coverage (each bp
sequenced on average 100 times) currently
costs less than 250 euro per genome on an
Illumina NextSeq500 machine (if 35 genomes
are pooled onto one lane). This is a stark difference with the multi-million euro Sanger
sequencing efforts of the past (Goffeau et al.
1996).
Since the early 2010s, new technologies
have become commercially available that produce considerably longer reads than Illumina.
Pacific Biosciences (PacBio) is based on singlemolecule sequencing and can produce reads of
on average 5 kbp and a maximum of 20 kbp
(Eid et al. 2009). Oxford Nanopore further
revolutionized sequencing by vastly reducing
the size of the machine to a mere USB flash
drive (Jain et al. 2016). This MinION machine
produces reads of over 100 kbp. However, both
208
R. A. Ohm
reads. Next, the genome assembly is annotated.
This entails the identification of repetitive
sequences, genes, and other functional elements in the genome. The predicted genes are
subsequently annotated by assigning a putative
function, usually based on homology to known
genes and domains.
This chapter will describe advances in technologies underlying fungal genome sequencing,
annotation, and analysis. Furthermore, the
impact of fungal genome sequencing is illustrated using examples from several fields of
biotechnology.
II. Advances in Genome Sequencing
Technologies
In the past decade, sequencing technologies
have improved dramatically, radically changing
the landscape of fungal genome sequencing.
Sanger sequencing was the first sequencing
technique that was used for genome sequencing
(Sanger et al. 1977b). It was used to sequence
landmark genomes such as the first bacteriophage jX174 (Sanger et al. 1977a), the first
bacterium Haemophilus influenza (Fleischmann et al. 1995), the first eukaryote (and first
fungus) Saccharomyces cerevisiae (Goffeau
et al. 1996), the first plant Arabidopsis thaliana
(Arabidopsis Genome Initiative 2000), the first
animal Caenorhabditis elegans (The C. elegans
Sequencing Consortium 1998), as well as the
human genome (International Human Genome
Sequencing Consortium 2001). The sequencing
of these genomes was generally a multi-year
undertaking and was performed by large consortia of collaborating labs. Sequencing reads
that were obtained with Sanger technology were
relatively long (up to approximately 1500 bp)
and were relatively straightforward to assemble
using assembly software such as Jazz (Aparicio
et al. 2002) or Arachne (Batzoglou et al. 2002).
More recently, since the mid-2000s, several
new sequencing platforms were developed that
are collectively known as “next-generation
sequencing” (NGS). Initially, these techniques
included the now mostly defunct technologies
Roche 454 (Margulies et al. 2005), IonTorrent
(Life Technologies) and SOLiD (Applied Biosystems). Currently the most prominent short
read sequencing technology, however, is Illumina (Bennett 2004). Although the sequences
generated by Illumina technology were initially
too short for efficient genome sequencing (up
to 25 bp), this has increased to currently
2 Â 300 bp on an Illumina MiSeq machine.
New assembly approaches and software were
developed for these short reads, such as Velvet
(Zerbino and Birney 2008), ABySS (Simpson
et al. 2009), SOAPdenovo (Luo et al. 2012),
and SPAdes (Bankevich et al. 2012).
Rather paradoxically, the assemblies generated from early NGS techniques were not nearly
as good as the ones generated from Sanger
reads, with respect to assembly fragmentation.
Especially repetitive genomic regions (e.g., originating from transposable elements) were
challenging to assembly using short reads.
However, crucial advantages of NGS technologies are that they are considerably faster and
cheaper than Sanger sequencing (Ghurye and
Pop 2019). This meant that genome sequencing
became affordable to core facilities and even
individual researchers, as opposed to the large
sequencing consortia that were required for
Sanger-based genome sequencing. This is illustrated by the following back-of-the-envelope
calculation: sequencing a typical fungal genome
of 30 Mbp with 100-fold coverage (each bp
sequenced on average 100 times) currently
costs less than 250 euro per genome on an
Illumina NextSeq500 machine (if 35 genomes
are pooled onto one lane). This is a stark difference with the multi-million euro Sanger
sequencing efforts of the past (Goffeau et al.
1996).
Since the early 2010s, new technologies
have become commercially available that produce considerably longer reads than Illumina.
Pacific Biosciences (PacBio) is based on singlemolecule sequencing and can produce reads of
on average 5 kbp and a maximum of 20 kbp
(Eid et al. 2009). Oxford Nanopore further
revolutionized sequencing by vastly reducing
the size of the machine to a mere USB flash
drive (Jain et al. 2016). This MinION machine
produces reads of over 100 kbp. However, both
208
R. A. Ohm