This is a protocol preprint. Preprints are preliminary reports that have not undergone peer review. They should not be considered conclusive, used to inform clinical practice, or referenced by the media as validated information.
Method Article
A MAKER pipeline for prediction of protein-coding genes in parasitic worm genomes
Matthew Berriman
Berriman Lab Group, Sanger Institute
Corresponding Author
License:
This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License
Once a genome has been sequenced, a challenging task is to predict protein-coding genes in the genome assembly. If several genomes of related species have been sequenced, it is desirable that the pipeline is scalable to multiple species, and will give consistent results across species. Here we describe a consistent, scalable and automated computational pipeline, based on the MAKER software, for prediction of protein-coding genes in parasitic worm (nematode and platyhelminth) genome assemblies. This protocol can be used to generate a relatively accurate and complete gene set, for a broad range of nematode and platyhelminth species. Furthermore, it does not require RNAseq data and does not require training using a curated set of genes of known structure.
Keywords
MAKER, Augustus, GeneMark, SNAP, RATT, genBlastG, gene-finding, annotation, gene prediction, gene models, gene set, parasites, genomes, assemblies, parasitic worms, helminths, nematodes, flatworms, platyhelminths, Nematoda, Platyhelminthes
Figure 1
Once a genome has been sequenced, a challenging task is to predict protein-coding genes in the genome assembly. If several genomes of related species have been sequenced, it is desirable that the pipeline is scalable to multiple species, and will give consistent results across species. Here we describe a consistent, scalable and automated computational pipeline, based on the MAKER software, for prediction of protein-coding genes in parasitic worm (nematode and platyhelminth) genome assemblies.
In our pipeline, gene predictions are generated using a pipeline that uses MAKER version 2.2.28
(Holt et al 2011). Our MAKER annotation pipeline consists of four steps, taking into account evidence from multiple sources (Figure 1). First, repetitive elements in the genome are identified and masked using RepeatMasker (www.repeatmasker.org) by scanning scaffolds for matches to repeats from a repeat library generated using RepeatModeler (www.repeatmasker.org/RepeatModeler.html). Second, ab initio gene models to be used as evidence within MAKER are generated using Augustus 2.5.5 (Stanke et al 2006), GeneMark-ES 2.3a (self-trained) (Ter-Hovhannisyan et al 2008), and SNAP 2013-02-16 (Korf 2004). Further gene models to use as MAKER input are generated using comparative algorithms genBlastG (She et al 2011) (using comparisons to C. elegans gene models from WormBase; Yook et al 2012) and RATT (Otto et al 2011; transferring gene models from the taxonomically nearest published ‘reference’ genome from the list: Haemonchus contortus for clade V parasites; Ascaris suum for Ascaridomorpha; Brugia malayi (and Onchocerca volvulus) for Spiruromorpha; Trichuris muris for clade I; Strongyloides ratti for clade IV; Hymenolepis microstoma for cestodes except Echinococcus multilocularis for Taenia species; Schistosoma mansoni for trematodes). Third, species-specific ESTs and cDNAs from INSDC (Cochrane et al 2016), and proteins from related species (see below), are aligned against the genome using BLASTN and BLASTX (Altschul et al 1997), respectively, and these alignments are further refined with respect to splice sites using exonerate (Slater & Birney 2005). Last, the EST and protein homology alignments, comparative gene models, and ab initio gene predictions are integrated and filtered by MAKER to produce a gene set for the species, with just one transcript for each gene.
The four-step MAKER pipeline is run three consecutive times (Figure 1). The first run is performed using the est2genome option with species-specific ESTs and cDNAs and the protein2genome for nematode protein sequences from UniProt’s UniRef 90 clusters for nematodes (UniProt Consortium, 2015). For this first MAKER run, Augustus and SNAP are trained using CEGMA (Parra et al 2007) gene models for KOGs, as well as ‘nematode orthologous groups’ (NOGs) (Martin & Mitreva 2018), ‘trematode orthologous groups’ (TROGs), or ‘cestode orthologous groups’ (CEOGs) as appropriate.
Gene models obtained from the first MAKER run are used to train SNAP, and MAKER is run a second time, using the same nematode proteins as in the first run. Gene models from the second run are then used to train Augustus. Using the trained versions of SNAP and Augustus, MAKER is run a third time, using a taxonomically broader protein set that includes proteins from metazoans with complete proteomes from UniProt and proteins from helminths from GeneDB (Logan-Klumpler et al 2012). The resulting MAKER gene set is then filtered to remove less reliable gene models.
This protocol can be used to generate a relatively accurate and complete gene set, for a broad range of nematode and platyhelminth species. Unlike gene-finding approaches that require extensive use of RNAseq data (for example, use of Augustus in the Strongyloides project; Hunt et al 2016), this protocol does not require RNAseq data, so is suitable when RNAseq data is not available. Furthermore, unlike many gene-finding approaches, it does not require training using a curated set of genes of known structure (e.g. confirmed using cDNAs or RNAseq), as it uses conserved genes from KOGs/NOGs/TROGs/CEOGs for an initial round of training, and then its own gene MAKER models for later rounds of training.
Computer cluster.
Gene predictions are generated using MAKER version 2.2.28 (Holt et al 2011). The MAKER annotation pipeline consists of four steps, taking into account evidence from multiple sources. The four-step MAKER pipeline is run three consecutive times ('round 1', 'round 2', 'round 3' in Figure 1).
Step 1
Generate a repeat library for the genome of interest using RepeatModeler (www.repeatmasker.org/RepeatModeler.html).
Identity and mask repetitive elements in the genome using RepeatMasker (www.repeatmasker.org) by scanning scaffolds for matches to the repeats in the repeat library.
Step 2
For the first MAKER run ('round 1' in Figure 1), train SNAP 2013-02-16 (Korf 2004) and Augustus 2.5.5 (Stanke et al 2006) using CEGMA (Parra et al 2007) gene models for KOGs, as well as ‘nematode orthologous groups’ (NOGs) (Martin & Mitreva 2018), ‘trematode orthologous groups’ (TROGs), or ‘cestode orthologous groups’ (CEOGs) as appropriate.
For the second MAKER run ('round 2'), first train SNAP using the MAKER gene models obtained from the first MAKER run ('round 1'). For the third MAKER run ('round 3'), first train Augustus using the MAKER gene models obtained from the second MAKER run ('round 2').
Generate ab initio gene models to be used as evidence within MAKER, using Augustus, GeneMark-ES 2.3a (self-trained) (Ter-Hovhannisyan et al 2008), and SNAP.
Generate further gene models to use as MAKER input, using comparative algorithms genBlastG (She et al 2011) (using comparisons to C. elegans gene models from WormBase; Yook et al 2012) and RATT (Otto et al 2011; transferring gene models from the taxonomically nearest published ‘reference’ genome from the list: Haemonchus contortus for clade V parasites; Ascaris suum for Ascaridomorpha; Brugia malayi (and Onchocerca volvulus) for Spiruromorpha; Trichuris muris for clade I; Strongyloides ratti for clade IV; Hymenolepis microstoma for cestodes except Echinococcus multilocularis for Taenia species; Schistosoma mansoni for trematodes).
Step 3
Align species-specific ESTs and cDNAs from INSDC (Cochrane et al 2016), and proteins from related species (see below), against the genomes using BLASTN and BLASTX (Altschul et al 1997), respectively.
The first and second MAKER runs ('round 1' and 'round 2' in Figure 1) are performed using the est2genome option with species-specific ESTs and cDNAs, and the protein2genome for nematode protein sequences from UniProt’s UniRef 90 clusters for nematodes (UniProt Consortium, 2015). The third MAKER run ('round 3') is performed using species-specific ESTs and cDNAs, and a taxonomically broader protein set that includes proteins from metazoans with complete proteomes from UniProt and proteins from helminths from GeneDB (Logan-Klumpler et al 2012).
Refine these alignments further with respect to splice sites using exonerate (Slater & Birney 2005).
Step 4
The EST and protein homology alignments, comparative gene models, and ab initio gene predictions are integrated and filtered by MAKER to produce a gene set for the species, with just one transcript for each gene.
Step 5
After running the MAKER pipeline above (Steps 1-4) three times (rounds 1-3), filter the resulting MAKER gene set to remove less reliable gene models, as follows:
Discard any MAKER gene models that were based on exonerate or BLASTX alignments, and do not overlap any Augustus, genBlastG or RATT gene model (as they were probably due to spurious alignments).
Discard MAKER gene models that encode proteins of shorter than 30 amino acids.
- If two different MAKER gene models overlap in their coding sequence, discard the gene model with the worst MAKER score (i.e. AED score).
It is likely that in a highly fragmented genome assembly, many genes will be split into multiple partial gene predictions. To correct for the effect of assembly fragmentation on the gene count, the gene count in your assembly can be normalised, by dividing the total proteome length by the mean protein length for C. elegans (409.82 amino acids). This should give a more accurate estimate of the true gene count in your species.
The output from the protocol is a set of protein-coding gene predictions for one or more genome assemblies of interest.
Altschul, S. F. et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic acids research 25, 3389-3402 (1997).
Cochrane, G. et al. The International Nucleotide Sequence Database Collaboration. Nucleic acids research 44, D48-50 (2016).
Holt, C. & Yandell, M. MAKER2: an annotation pipeline and genome-database management tool for second-generation genome projects. BMC bioinformatics 12, 491 (2011).
Hunt, V. L. et al. The genomic basis of parasitism in the Strongyloides clade of nematodes. Nat Genet. 48, 299-307 (2016).
Korf, I. Gene finding in novel genomes. BMC bioinformatics 5, 59 (2004).
Logan-Klumpler, F. J. et al. GeneDB--an annotation database for pathogens. Nucleic acids research 40, D98-108 (2012).
Martin, J. & Mitreva, M. Finding set of "core genes" for Nematoda, Trematoda and Cestoda. Protocol Exchange (2018).
Otto, T. D., et al. RATT: Rapid Annotation Transfer Tool. Nucleic acids research 39, e57 (2011).
Parra, G. et al. CEGMA: a pipeline to accurately annotate core genes in eukaryotic genomes. Bioinformatics 23, 1061-1067 (2007).
She, R. et al. genBlastG: using BLAST searches to build homologous gene models. Bioinformatics 27, 2141-2143 (2011).
Slater, G. S. & Birney, E. Automated generation of heuristics for biological sequence comparison. BMC bioinformatics 6, 31 (2005).
Stanke, M. et al. AUGUSTUS: ab initio prediction of alternative transcripts. Nucleic acids research 34, W435-439 (2006).
Ter-Hovhannisyan, V. et al. Gene prediction in novel fungal genomes using an ab initio algorithm with unsupervised training. Genome Res. 18, 1979-1990 (2008).
UniProt Consortium. UniProt: a hub for protein information. Nucleic acids research 43, D204-212 (2015).
Yook, K. et al. WormBase 2012: more genomes, more data, new website. Nucleic acids research 40, D735-741 (2012).
We would like to thank the WSI Pathogen Informatics team, especially Jacqueline Keane and Andrew Page.
Version History
CURRENT STATUS: Posted
Version 1
Posted 14 May, 2018
Metrics
Comments: 0
HTML Views: ...
Method Article
A MAKER pipeline for prediction of protein-coding genes in parasitic worm genomes
Matthew Berriman
Berriman Lab Group, Sanger Institute
Corresponding Author
Version History
CURRENT STATUS: Posted
Version 1
Posted 14 May, 2018
Metrics
Comments: 0
HTML Views: ...
License:
This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License
Once a genome has been sequenced, a challenging task is to predict protein-coding genes in the genome assembly. If several genomes of related species have been sequenced, it is desirable that the pipeline is scalable to multiple species, and will give consistent results across species. Here we describe a consistent, scalable and automated computational pipeline, based on the MAKER software, for prediction of protein-coding genes in parasitic worm (nematode and platyhelminth) genome assemblies. This protocol can be used to generate a relatively accurate and complete gene set, for a broad range of nematode and platyhelminth species. Furthermore, it does not require RNAseq data and does not require training using a curated set of genes of known structure.
Figure 1
Once a genome has been sequenced, a challenging task is to predict protein-coding genes in the genome assembly. If several genomes of related species have been sequenced, it is desirable that the pipeline is scalable to multiple species, and will give consistent results across species. Here we describe a consistent, scalable and automated computational pipeline, based on the MAKER software, for prediction of protein-coding genes in parasitic worm (nematode and platyhelminth) genome assemblies.
In our pipeline, gene predictions are generated using a pipeline that uses MAKER version 2.2.28
(Holt et al 2011). Our MAKER annotation pipeline consists of four steps, taking into account evidence from multiple sources (Figure 1). First, repetitive elements in the genome are identified and masked using RepeatMasker (www.repeatmasker.org) by scanning scaffolds for matches to repeats from a repeat library generated using RepeatModeler (www.repeatmasker.org/RepeatModeler.html). Second, ab initio gene models to be used as evidence within MAKER are generated using Augustus 2.5.5 (Stanke et al 2006), GeneMark-ES 2.3a (self-trained) (Ter-Hovhannisyan et al 2008), and SNAP 2013-02-16 (Korf 2004). Further gene models to use as MAKER input are generated using comparative algorithms genBlastG (She et al 2011) (using comparisons to C. elegans gene models from WormBase; Yook et al 2012) and RATT (Otto et al 2011; transferring gene models from the taxonomically nearest published ‘reference’ genome from the list: Haemonchus contortus for clade V parasites; Ascaris suum for Ascaridomorpha; Brugia malayi (and Onchocerca volvulus) for Spiruromorpha; Trichuris muris for clade I; Strongyloides ratti for clade IV; Hymenolepis microstoma for cestodes except Echinococcus multilocularis for Taenia species; Schistosoma mansoni for trematodes). Third, species-specific ESTs and cDNAs from INSDC (Cochrane et al 2016), and proteins from related species (see below), are aligned against the genome using BLASTN and BLASTX (Altschul et al 1997), respectively, and these alignments are further refined with respect to splice sites using exonerate (Slater & Birney 2005). Last, the EST and protein homology alignments, comparative gene models, and ab initio gene predictions are integrated and filtered by MAKER to produce a gene set for the species, with just one transcript for each gene.
The four-step MAKER pipeline is run three consecutive times (Figure 1). The first run is performed using the est2genome option with species-specific ESTs and cDNAs and the protein2genome for nematode protein sequences from UniProt’s UniRef 90 clusters for nematodes (UniProt Consortium, 2015). For this first MAKER run, Augustus and SNAP are trained using CEGMA (Parra et al 2007) gene models for KOGs, as well as ‘nematode orthologous groups’ (NOGs) (Martin & Mitreva 2018), ‘trematode orthologous groups’ (TROGs), or ‘cestode orthologous groups’ (CEOGs) as appropriate.
Gene models obtained from the first MAKER run are used to train SNAP, and MAKER is run a second time, using the same nematode proteins as in the first run. Gene models from the second run are then used to train Augustus. Using the trained versions of SNAP and Augustus, MAKER is run a third time, using a taxonomically broader protein set that includes proteins from metazoans with complete proteomes from UniProt and proteins from helminths from GeneDB (Logan-Klumpler et al 2012). The resulting MAKER gene set is then filtered to remove less reliable gene models.
This protocol can be used to generate a relatively accurate and complete gene set, for a broad range of nematode and platyhelminth species. Unlike gene-finding approaches that require extensive use of RNAseq data (for example, use of Augustus in the Strongyloides project; Hunt et al 2016), this protocol does not require RNAseq data, so is suitable when RNAseq data is not available. Furthermore, unlike many gene-finding approaches, it does not require training using a curated set of genes of known structure (e.g. confirmed using cDNAs or RNAseq), as it uses conserved genes from KOGs/NOGs/TROGs/CEOGs for an initial round of training, and then its own gene MAKER models for later rounds of training.
Computer cluster.
Gene predictions are generated using MAKER version 2.2.28 (Holt et al 2011). The MAKER annotation pipeline consists of four steps, taking into account evidence from multiple sources. The four-step MAKER pipeline is run three consecutive times ('round 1', 'round 2', 'round 3' in Figure 1).
Step 1
Generate a repeat library for the genome of interest using RepeatModeler (www.repeatmasker.org/RepeatModeler.html).
Identity and mask repetitive elements in the genome using RepeatMasker (www.repeatmasker.org) by scanning scaffolds for matches to the repeats in the repeat library.
Step 2
For the first MAKER run ('round 1' in Figure 1), train SNAP 2013-02-16 (Korf 2004) and Augustus 2.5.5 (Stanke et al 2006) using CEGMA (Parra et al 2007) gene models for KOGs, as well as ‘nematode orthologous groups’ (NOGs) (Martin & Mitreva 2018), ‘trematode orthologous groups’ (TROGs), or ‘cestode orthologous groups’ (CEOGs) as appropriate.
For the second MAKER run ('round 2'), first train SNAP using the MAKER gene models obtained from the first MAKER run ('round 1'). For the third MAKER run ('round 3'), first train Augustus using the MAKER gene models obtained from the second MAKER run ('round 2').
Generate ab initio gene models to be used as evidence within MAKER, using Augustus, GeneMark-ES 2.3a (self-trained) (Ter-Hovhannisyan et al 2008), and SNAP.
Generate further gene models to use as MAKER input, using comparative algorithms genBlastG (She et al 2011) (using comparisons to C. elegans gene models from WormBase; Yook et al 2012) and RATT (Otto et al 2011; transferring gene models from the taxonomically nearest published ‘reference’ genome from the list: Haemonchus contortus for clade V parasites; Ascaris suum for Ascaridomorpha; Brugia malayi (and Onchocerca volvulus) for Spiruromorpha; Trichuris muris for clade I; Strongyloides ratti for clade IV; Hymenolepis microstoma for cestodes except Echinococcus multilocularis for Taenia species; Schistosoma mansoni for trematodes).
Step 3
Align species-specific ESTs and cDNAs from INSDC (Cochrane et al 2016), and proteins from related species (see below), against the genomes using BLASTN and BLASTX (Altschul et al 1997), respectively.
The first and second MAKER runs ('round 1' and 'round 2' in Figure 1) are performed using the est2genome option with species-specific ESTs and cDNAs, and the protein2genome for nematode protein sequences from UniProt’s UniRef 90 clusters for nematodes (UniProt Consortium, 2015). The third MAKER run ('round 3') is performed using species-specific ESTs and cDNAs, and a taxonomically broader protein set that includes proteins from metazoans with complete proteomes from UniProt and proteins from helminths from GeneDB (Logan-Klumpler et al 2012).
Refine these alignments further with respect to splice sites using exonerate (Slater & Birney 2005).
Step 4
The EST and protein homology alignments, comparative gene models, and ab initio gene predictions are integrated and filtered by MAKER to produce a gene set for the species, with just one transcript for each gene.
Step 5
After running the MAKER pipeline above (Steps 1-4) three times (rounds 1-3), filter the resulting MAKER gene set to remove less reliable gene models, as follows:
Discard any MAKER gene models that were based on exonerate or BLASTX alignments, and do not overlap any Augustus, genBlastG or RATT gene model (as they were probably due to spurious alignments).
Discard MAKER gene models that encode proteins of shorter than 30 amino acids.
- If two different MAKER gene models overlap in their coding sequence, discard the gene model with the worst MAKER score (i.e. AED score).
It is likely that in a highly fragmented genome assembly, many genes will be split into multiple partial gene predictions. To correct for the effect of assembly fragmentation on the gene count, the gene count in your assembly can be normalised, by dividing the total proteome length by the mean protein length for C. elegans (409.82 amino acids). This should give a more accurate estimate of the true gene count in your species.
The output from the protocol is a set of protein-coding gene predictions for one or more genome assemblies of interest.
Altschul, S. F. et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic acids research 25, 3389-3402 (1997).
Cochrane, G. et al. The International Nucleotide Sequence Database Collaboration. Nucleic acids research 44, D48-50 (2016).
Holt, C. & Yandell, M. MAKER2: an annotation pipeline and genome-database management tool for second-generation genome projects. BMC bioinformatics 12, 491 (2011).
Hunt, V. L. et al. The genomic basis of parasitism in the Strongyloides clade of nematodes. Nat Genet. 48, 299-307 (2016).
Korf, I. Gene finding in novel genomes. BMC bioinformatics 5, 59 (2004).
Logan-Klumpler, F. J. et al. GeneDB--an annotation database for pathogens. Nucleic acids research 40, D98-108 (2012).
Martin, J. & Mitreva, M. Finding set of "core genes" for Nematoda, Trematoda and Cestoda. Protocol Exchange (2018).
Otto, T. D., et al. RATT: Rapid Annotation Transfer Tool. Nucleic acids research 39, e57 (2011).
Parra, G. et al. CEGMA: a pipeline to accurately annotate core genes in eukaryotic genomes. Bioinformatics 23, 1061-1067 (2007).
She, R. et al. genBlastG: using BLAST searches to build homologous gene models. Bioinformatics 27, 2141-2143 (2011).
Slater, G. S. & Birney, E. Automated generation of heuristics for biological sequence comparison. BMC bioinformatics 6, 31 (2005).
Stanke, M. et al. AUGUSTUS: ab initio prediction of alternative transcripts. Nucleic acids research 34, W435-439 (2006).
Ter-Hovhannisyan, V. et al. Gene prediction in novel fungal genomes using an ab initio algorithm with unsupervised training. Genome Res. 18, 1979-1990 (2008).
UniProt Consortium. UniProt: a hub for protein information. Nucleic acids research 43, D204-212 (2015).
Yook, K. et al. WormBase 2012: more genomes, more data, new website. Nucleic acids research 40, D735-741 (2012).
We would like to thank the WSI Pathogen Informatics team, especially Jacqueline Keane and Andrew Page.