The Vertebrate Genome Project (VGP), a project of the Genome 10K (G10K) Consortium, aims to generate high-quality, near error-free, gap-free, chromosome-level, haplotype-phased, annotated reference genome assemblies for every vertebrate species (Rhie et al. 2020). The VGP has developed a fully automated de-novo genome assembly pipeline, which uses a combination of three different technologies: Pacbio high fidelity reads (HiFi), all-versus-all chromatin conformation capture (Hi-C) data, and (optionally) BioNano optical map data. The pipeline consists of nine distinct workflows. This tutorial provides a quick example of how to run these workflows for one particular scenario, which is, based on our experience, the most common: assembling genomes using HiFi Reads combined with Hi-C data (both generated from the same individual).
This tutorial assumes you are comfortable getting data into Galaxy, running jobs, managing history, etc. If you are unfamiliar with Galaxy, we recommend you visit the Galaxy Training Network. Consider starting with the following trainings:
Better haplotype resolution and improved continuity
F
HiFi + parental + BioNano
Better haplotype resolution and improved continuity
G
HiFi + parental data + Hi-C + BioNano
Better haplotype resolution and ultimate continuity
H
If this table “HiFi” and “Hi-C” are derived from the individual whose genome is being assembled. “Parental data” is high coverage Illumina data derived from parents of the individual being assembled. Datasets containing parental data are also called “Trios”. Each combination of input datasets is supported by an analysis trajectory: a combination of workflows designed for generating assembly given a particular combination of inputs. These trajectories are listed in the table above and shown in the figure below. We suggest at least 30✕ PacBio HiFi coverage and 30✕ Hi-C coverage per haplotype (parental genome); and up to 60✕ coverage to accurately assemble highly repetitive regions.
Figure 1: Eight analysis trajectories are possible depending on the combination of input data. A decision on whether or not to invoke Workflow 6 is based on the analysis of QC output of workflows 3, 4, or 5. Thicker lines connecting Workflows 7, 8, and 9 represent the fact that these workflows are invoked separately for each phased assembly (once for maternal and once for paternal).
The first stage of the pipeline is the generation of k-mer profiles of the raw reads to estimate genome size, heterozygosity, repetitiveness, and error rate necessary for parameterizing downstream workflows. The generation of k-mer counts can be done from HiFi data only (Workflow 1) or include data from parental reads for trio-based phasing (Workflow 2; trio is a combination of paternal sequencing data with that from an offspring that is being assembled). The second stage is the phased contig assembly. In addition to using only HiFi reads (Workflow 3), the contig building (contiging) step can leverage Hi-C (Workflow 4) or parental read data (Workflow 5) to produce fully-phased haplotypes (hap1/hap2 or parental/maternal assigned haplotypes), using hifiasm. The contiging workflows also produce a number of critical quality control (QC) metrics such as k-mer multiplicity profiles. Inspection of these profiles provides information to decide whether the third stage—purging of false duplication—is required. Purging (Workflow 6), using purge_dups identifies and resolves haplotype-specific assembly segments incorrectly labeled as primary contigs, as well as heterozygous contig overlaps. This increases continuity and the quality of the final assembly. The purging stage is generally unnecessary for trio data for which reliable haplotype resolution is performed using k-mer profiles obtained from parental reads. The fourth stage, scaffolding, produces chromosome-level scaffolds using information provided by Bionano (Workflow 7), with Bionano Solve (optional) and Hi-C (Workflow 8) data and YaHS algorithms. A final stage of decontamination (Workflow 9) removes exogenous sequences (e.g., viral and bacterial sequences) from the scaffolded assembly. A separate workflow (WF0) is used for mitochondrial assembly.
Comment: A note on data quality
We suggest at least 30✕ PacBio HiFi coverage and 30✕ Hi-C coverage per haplotype (parental genome); and up to 60✕ coverage to accurately assemble highly repetitive regions.
Getting the data
The following steps use PacBio HiFi and Illumina Hi-C data from baker’s yeast (Saccharomyces cerevisiae). The tutorial represents trajectory B from Fig. 1 above. For this tutorial, the first step is to get the datasets from Zenodo. Specifically, we will be uploading two datasets:
A set of PacBio HiFi reads in fasta format
A set of Illumina Hi-C reads in fastqsanger.gz format
Uploading fasta datasets from Zenodo
The following two steps demonstrate how to upload three PacBio HiFi datasets into you Galaxy history.
Hands-on: Uploading FASTA datasets from Zenodo
Create a new history for this tutorial
Click the new-history icon at the top of the history panel:
Copy the following URLs into clipboard.
you can do this by clicking on copy button in the right upper corner of the box below. It will appear if you mouse over the box.)
Click galaxy-uploadUpload Data at the top of the tool panel
Select galaxy-wf-editPaste/Fetch Data
Paste the link(s) into the text field
Change Type (set all): from “Auto-detect” to fasta
Press Start
Close the window
Uploading fastqsanger or fastqsanger.gz datasets via URL.
Click on Upload Data on the top of the left panel:
Click on Paste/Fetch:
Paste URL into text box that would appear:
Set Type (set all) to fastqsanger or, if your data is compressed as in URLs above (they have .gz extensions), to fastqsanger.gz
:
Warning: Danger: Make sure you choose corect format!
When selecting datatype in “Type (set all)” dropdown, make sure you select fastaqsanger or fastqsanger.gz BUT NOT fastqcssanger or anything else!
Warning: These datasets are large!
Hi-C datasets are large. It will take some time (~15 min) for them to be fully uploaded. Please, be patient.
Organizing the data
If everything goes smoothly you history will look like shown in Fig. 4 below. The three HiFi fasta files are better represented as a collection: Galaxy’s way to represent multiple datasets as a single interface entity (collection). Also, importantly, the workflow we will be using for the analysis of our data takes collection as an input (it does not access individual datasets). So let’s create a collection using steps outlines in the Tip tip “Creating a dataset collection” that you can find below Fig. 4.
Figure 2: History after uploading HiFi and HiC data (left). Creation of a list (collection) combines all HiFi datasets into a single history item called 'HiFi data' (right). See below for instruction on how to make this collection.
Click on galaxy-selectorSelect Items at the top of the history panel
Check all the datasets in your history you would like to include
Click n of N selected and choose Build Dataset List
Enter a name for your collection
Click Create List to build your collection
Click on the checkmark icon at the top of your history again
You can obviously upload your own datasets via URLs as illustrated above or from your own computer. In addition, you can upload data from a major repository called GenomeArk. GenomeArk is integrated directly into Galaxy Upload. To use GenomeArk following the steps in the Tip tip below:
Open the file galaxy-uploadupload menu
Click on Choose remote files tab
Click on the Genome Ark button and then click on species
You can find the data by following this path: /species/${Genus}_${species}/${specimen_code}/genomic_data. Inside a given datatype directory (e.g.pacbio), select all the relevant files individually until all the desired files are highlighted and click the Ok button. Note that there may be multiple pages of files listed. Also note that you may not want every file listed.
Once we have imported the datasets, the next step is to import the workflows necessary for the analysis of our data from DockStore.
Importing workflows
All analyses described in this tutorial are performed using workflows–chains of tools–shown in Fig. 1. Specifically, we will use four workflows corresponding to analysis trajectory B: 1, 4, 6, and 8. To use these four workflows you need to import them into your Galaxy account following the steps below:
Hands-on: Importing workflows from GitHub
Links to the four workflows that will be used in this tutorial are listed in the table. Follow the procedure described below the table to import each of them into your Galaxy account.
Click on “Galaxy” dropdown within the “Launch with” panel located in the upper right corner.
Select a galaxy instance you want to launch this workflow with.
You will be redirected to Galaxy and presented with a list of workflow versions.
Click the version you want (usually the latest labelled as “main”)
You are done!
Warning: Make sure you are logged in!
Ensure that you are logged in into your Galaxy account!
The following short video walks you through this uncomplicated procedure:
WorkflowHub is a workflow management system which allows workflows to be FAIR (Findable, Accessible, Interoperable, and Reusable), citable, have managed metadata profiles, and be openly available for review and analytics.
Warning: Make sure you are logged in!
Ensure that you are logged in into your Galaxy account!
Click on the Workflow menu, located in the top bar.
Click on the Import button, located in the right corner.
In the section “Import a Workflow from Configured GA4GH Tool Registry Servers (e.g. Dockstore)”, click on Search form.
In the TRS Server: workflowhub.eu menu you should type your query.
Click on the desired workflow, and finally select the latest available version.
After that, the imported workflows will appear in the main workflow menu. In order to run the workflow, just need to click in the workflow-runRun workflow icon.
Below is a short video showing this uncomplicated procedure:
Click on Workflow on the top menu bar of Galaxy. You will see a list of all your workflows.
Click on galaxy-uploadImport at the top-right of the screen
Provide your workflow
Option 1: Paste the URL of the workflow into the box labelled “Archived Workflow URL”
Option 2: Upload the workflow file in the box labelled “Archived Workflow File”
Click the Import workflow button
Below is a short video demonstrating how to import a workflow from GitHub using this procedure:
Once all four workflows are imported, your workflow list should look like this:
Figure 4: Workflow list. The workflow menu lists all the workflows that have been imported. It provides useful information for organizing the workflows, such as last update and the tags. The workflows can be run by clicking in the play icon, marked in red in the image.
Once we have imported the datasets and the workflows, we can start with the genome assembly.
Performing the assembly
Workflows listed in Fig. 1 support a variety of “analysis trajectories”. The majority of species that were sequenced by the VGP usually contain HiFi reads for the individual being sequenced supplemented with Hi-C data. As a result most assemblies performed by us follow the trajectory B. This is why this tutorial was designed to follow this trajectory as well.
Genome profile analysis (WF1)
Now that our data and workflows are imported, we can run our first workflow. Before the assembly can be run, we need to collect metrics on the properties of the genome under consideration, such as the expected genome size according to our data. The present pipeline uses Meryl for generating the k-mer database and Genomescope2 for determining genome characteristics based on a k-mer analysis.
Figure 5: Workflow main menu. The workflow menu lists all the workflows that have been imported. It provides useful information for organizing the workflows, such as last update and the tags. The worklows can be run by clicking in the play icon, marked in red in the image.
Comment: K-mer length
In this tutorial, we are using a k-mer length of 31. This can vary, but the VGP pipeline tends to use a k-mer length of 21, which tends to work well for most mammalian-size genomes. There is more discussion about k-mer length trade-offs in the extended VGP pipeline tutorial.
Refill your coffee
Assembly is not exactly an instantaneous type of analysis - this workflow will take approx 15 minutes to complete. The same is true for all analyses in tutorial.
Interpreting the results
Once the workflow has finished, we can evaluate the linear plot generated by Genomescope, which includes valuable information such as the observed k-mer profile, fitted models and estimated parameters. This file corresponds to the dataset 15 in this history.
Figure 6: GenomeScope2 k-mer profile. The first peak located at about 25× corresponds to the heterozygous peak. The second peak at 50×, corresponds to the homozygous peak. The plot also includes information about the the inferred total genome length (len), genome unique length percent (uniq), overall heterozygosity rate (ab), mean k-mer coverage for heterozygous bases (kcov), read error rate (err), average rate of read duplications (dup) and k-mer size (k).
This distribution is the result of the Poisson process underlying the generation of sequencing reads. As we can see, the k-mer profile follows a bimodal distribution, indicative of a diploid genome. The distribution is consistent with the theoretical diploid model (model fit > 93%). Low frequency k-mers are the result of sequencing errors, and are indicated by the red line. Genomescope2 estimated a haploid genome size of around 11.7 Mbp, a value reasonably close to the Saccharomyces genome size.
Assembly (contiging) with hifiasm (WF4)
To generate contiguous sequences in an assembly (contigs) we will use hifiasm assembler. It is a part of the Assembly with HiC (WF4) workflow . This workflow uses hifiasm (HiC mode) to generate HiC-phased haplotypes (hap1 and hap2). This is in contrast to its default mode, which generates primary and alternate pseudohaplotype assemblies. This workflow includes three tools for evaluating assembly quality: gfastats, BUSCO and Merqury.
Launching the workflow
Hands-on: Launching assembly (contiging) workflow
Step 1: Identify inputs
The assembly workflow takes the following inputs:
HiFi reads as a collection
Forward Hi-C reads
Reverse Hi-C reads
Genomescope Model Parameters generated by previous (k-mer profiling) workflow
Genomescope Summary generated by previous (k-mer profiling) workflow
Meryl k-mer database generated by previous (k-mer profiling) workflow
Busco lineage
Step 2: Launch the workflow
Click in the Workflow menu, located in the top bar
Click in the workflow-runRun workflow button corresponding to VGP HiFi phased assembly with hifiasm and HiC data
In the Workflow: Assembly with HiC (WF4) menu fill the following parameters:
param-collection “Pacbio Reads Collection”: Collection with original HiFi data
param-file “Meryl database”: Meryl k-mer database: one of the outputs of the previous workflow (contains tag “MerylDatabase”)
param-file “Provide lineage for BUSCO (e.g., Vertebrata)”: Ascomycota
param-file “GenomeScope Summary”: GenomeScope summary: one of the outputs of the previous workflow (contains tag “GenomeScopeSummary”)
param-file “GenomeScope Model Parameters”: GenomeScope model parameters: one of the outputs of the previous workflow (contains tag “GenomeScopeParameters”)
Click on the Run workflow button
Interpreting the results
Warning: There will be two assemblies!
Because we are assembling a diploid organism this workflow will produce two assemblies: hap1 and hap2!
Let’s have a look at the stats generated by gfastats. This output summarizes some main assembly statistics, such as contig number, N50, assembly length, etc. Below we provide a partial output of gfastats in which information about both assemblies is shown side-by-side:
Statistic
Hap 1
Hap 2
# contigs
16
19
Total contig length
12,050,076
12,360,746
Average contig length
753,129.75
650,565.58
Contig N50
923,452
922,430
Contig N50
923,452
922,430
Contig auN
909,022.62
891,508.36
Contig L50
6
6
Contig L50
6
6
Contig NG50
923,452
922,430
Contig NG50
923,452
922,430
Contig auNG
932,462.97
938,074.26
Contig LG50
6
6
Contig LG50
6
6
Largest contig
1,532,843
1,531,728
Smallest contig
231,313
26,588
According to the report, both assemblies are quite similar; the primary assembly includes 16 contigs, whose cumulative length is around 12 Mbp. The alternate assembly includes 19 contigs, whose total length is 12.3Mbp. Both assemblies come close to the estimated genome size, which is as expected since we used hifiasm-HiC mode to generate phased assemblies which lowers the chance of false duplications that can inflate assembly size.
Comment: Are you working with pri/alt assemblies?
This tutorial uses the hifiasm-HiC workflow, which generates phased hap1 and hap2 assemblies. The phasing helps lower the chance of false duplications, since the phasing information helps the assembler know which genomic variation is heterozygosity at the same locus versus being two different loci entirely. If you are working with primary/alternate assemblies (especially if there is no internal purging in the initial assembly), you can expect higher false duplication rates than we observe here with the yeast HiC hap1/hap2.
Question
What is the longest contig in the primary assembly? And in the alternate one?
What is the N50 of the primary assembly?
The longest contig in the primary assembly is 1,532,843 bp, and 1,531,728 bp in the alternate assembly.
The N50 of the primary assembly is 923.452 bp.
Next, we are going to evaluate the outputs generated by BUSCO. This tool provides quantitative assessment of the completeness of a genome assembly in terms of expected gene content. It relies on the analysis of genes that should be present only once in a complete assembly or gene set, while allowing for rare gene duplications or losses (Simão et al. 2015).
Figure 7: A composite of BUSCO completeness summaries for hap1 and hap2
As we can see in the report, the results are simplified into four categories: complete and single-copy, complete and duplicated, fragmented and missing.
Question
How many complete BUSCO genes have been identified in the primary assembly?
How many BUSCOs genes are absent?
According to the report, our assembly contains the complete sequence of 1,562 complete BUSCO genes.
92 BUSCO genes are missing.
Despite BUSCO being robust for species that have been widely studied, it can be inaccurate when the newly assembled genome belongs to a taxonomic group that is not well represented in OrthoDB. Merqury provides a complementary approach for assessing genome assembly quality metrics in a reference-free manner via k-mer copy number analysis. Specifically, it takes our hap1 as the first genome assembly, hap2 as the second genome assembly, and the merylDB generated previously for k-mer counts.
By default, Merqury generates three collections as output: stats, plots and assembly consensus quality (QV) stats. The “stats” collection contains the completeness statistics, while the “QV stats” collection contains the quality value statistics. Let’s have a look at the copy number (CN) spectrum plot, known as the spectra-cn plot. The spectra-cn plot looks at both of your assemblies (here, your haplotypes) taken together (fig. 6a). We can see a small amount of false duplications here: at the 50 mark on the x-axis, there is a small amount of k-mers present at 3-copy across the two assemblies (the green bump).
Figure 8: Merqury CN plot for yeast assemblies. The plot tracks the multiplicity of each k-mer found in the read set and colors it by the number of times it is found in a given assembly. Merqury connects the midpoint of each histogram bin with a line, giving the illusion of a smooth curve. a). K-mer distribution of both haplotypes. b). K-mer distribution of an individual haplotype (hap2).
Thus, we know there is some false duplication (the 3-copy green bump) present as 2-copy in one of our assemblies, but we don’t know which one. We can look at the individual copy number spectrum for each haplotype in order to figure out which one contains the 2-copy k-mers (i.e., the false duplications). In the Merqury spectra-CN plot for hap2 we can see the small bump of 2-copy k-mers (blue) at around the 50 mark on the x-axis (fig. 6b).
Now that we know which haplotype contains the false duplications, we can run the purging workflow to try to get rid of these duplicates.
Purging duplicates with purge_dups
An ideal haploid representation would consist of one allelic copy of all heterozygous regions in the two haplotypes, as well as all hemizygous regions from both haplotypes (Guan et al. 2019). However, in highly heterozygous genomes, assembly algorithms are frequently not able to identify the highly divergent allelic sequences as belonging to the same region, resulting in the assembly of those regions as separate contigs. In order to prevent potential issues in downstream analysis, we are going to run the Purge duplicate contigs (WF6), which will allow to identify and reassign heterozygous contigs. This step is only necessary if haplotypic duplications are observed, and the output should be carefully checked for overpurging.
Launching the workflow
Hands-on: Launching duplicate purging workflow
Step 1: Identify inputs
The purging workflow takes the following inputs:
HiFi reads as a collection
Primary assembly produced by hifiasm in the previous run of assembly workflow (WF4).
Alternate assembly produced by hifiasam in the previous run of assembly workflow (WF4).
Genomescope Model Parameters generated by previous (k-mer profiling) workflow
Estimated genome size parsed from GenoeScope summary by the previous run of assembly workflow (WF4).
Meryl k-mer database generated by previous (k-mer profiling, WF1) workflow
Click in the Workflow menu, located in the top bar
Click in the workflow-runRun workflow button corresponding to Purge duplicate contigs (WF6)
In the Workflow: VGP purge assembly with purge_dups pipeline menu:
param-collection “Pacbio Reads Collection - Trimmed”: One of the outputs of the assembly workflow is a trimmed collection of HiFi reads. It has a tag trimmed_hifi.
param-file “Hifiasm Primary assembly”: An output of the assembly workflow (WF4) containing contigs for Hap1 in FASTA format. This dataset has a tag hifiasm_Assembly_Haplotype_1.
param-file “Hifiasm Alternate assembly”: An output of the assembly workflow (WF4) containing contigs for Hap2 in FASTA format. This dataset has a tag hifiasm_Assembly_Haplotype_2
param-file “Meryl database”: Meryl k-mer database: one of the outputs of the previous workflow (contains tag “MerylDatabase”)
param-file “GenomeScope Model Parameters”: GenomeScope model parameters: one of the outputs of the previous workflow (contains tag “GenomeScopeParameters”)
param-file “Estimated genome size”: A dataset produced with the assembly workflow (WF4). It contains a tag estimated_genome_size.
param-file “Provide lineage for BUSCO (e.g., Vertebrata)”: Ascomycota
Click in the Run workflow buttom
Interpreting results
The two most important outputs of the purging workflow are purged versions of Primary and Alternate assemblies. These have tags PurgedPrimaryAssembly and PurgedAlternateAssembly for Primary and Alternate assemblies, respectively. This step also provides QC metrics for evaluating the effect of purging (Figure below).
Figure 9: Comparison of pre- (a) and c)) and post-purging (b) and d)) Merqury CN spectra . The two top plots (a) and b)) for our dataset (yeast) and the two bottom plots (c) and d)) for a Chub mackerel (Scomber japonicus) -- a much larger genome. In the case of yeast the difference is not profound because our training dataset has been downsized and groomed to be as small as possible. In the case of zebra finch the green bump (k-mers appearing in three copies) is smaller after purging (Although potential overpurging can be seen by the new read-only (grey) bump that was not there before). Given the scale of the Y-axis this difference is substantial.
Hi-C scaffolding
In this final stage, we will run the Scaffolding HiC YAHS (WF8), which exploits the fact that the contact frequency between a pair of loci strongly correlates with the one-dimensional distance between them. This information allows YAHS – the main tool in this workflow – to generate scaffolds that are often chromosome-sized.
Launching Hi-C scaffolding workflow
Warning: The scaffolding workflow is run on ONE haplotype at a time.
Contiging (WF4) and purging (WF6) workflows work with both (hap1/hap2, primary/alternate) assemblies simultaneously. This is not the case for contiging – it hgas to be run independently for each haplotype assembly. In this example (below) we run contiging on hap1 (Primary) assembly only.
Hands-on: Launching Hi-C scaffolding workflow
Step 1: Identify inputs
The scaffolding workflow takes the following inputs:
An assembly graph
Forward Hi-C reads
Reverse Hi-C reads
Estimated genome size parsed from GenoeScope summary by the previous run of assembly workflow (WF4).
Restriction enzymes used in Hi-C library preparation procedure
Busco lineage
Step 2: Launch scaffolding workflow (WF8)
Click in the Workflow menu, located in the top bar
Click in the workflow-runRun workflow button corresponding to Scaffolding HiC YAHS (WF8)
In the Scaffolding HiC YAHS (WF8) menu:
param-file “input GFA”: Output of purging workflow (WF6) with a tag PurgedPrimaryAssembly (or PurgedPrimaryAssembly of scaffolding the Alternate assembly).
param-file “Estimated genome size - Parameter File”: An output of the contiging workflow (WF4) with a tag estimated_genome_size.
param-file “Provide lineage for BUSCO (e.g., Vertebrata)”: Ascomycota
Click in the Run workflow button
Comment: Bypassing purging workflow
In some situations (such as assemblies utilizing Trio data (Fig. 1) you do not need to perform purging and can go directly from contiging to scaffolding. In this case you will need to use an output of contiging workflow that has a tag hic_hap1_gfa for primary assembly or hic_hap2_gfa for alternate assembly:
In other words, the only parameter that you will need to set differently (relative to setting above) is this:
param-file “input GFA”: Output of contiging workflow (WF4) with a tag hic_hap1_gfa for primary assembly or hic_hap2_gfa for alternate assembly.
Interpreting the results
In order to evaluate the Hi-C hybrid scaffolding, we are going to compare the contact maps before and after running the HiC hybrid scaffolding workflow (Fig. below). They will have the following tags:
Before scaffolding: pretext_s1
After scaffolding: pretext_s2
Below is the comparison of the two maps obtained from our data a more profound “real live” example from assembly of zebra finch (Taeniopygia guttata) genome:
Figure 10: Hi-C maps generated by Pretext before and after scaffolding with Hi-C data. The red circles indicate the differences between the contact maps generated before and after Hi-C hybrid scaffolding. The bottom two panels show results of scaffolding on zebra finch where scaffolding dramatically decreases the number of segments by merging multiple contigs into scaffolds.
The regions marked with red circles highlight the most notable difference between the two contact maps, where inversion has been fixed.
Figure 11: Cumulative continuity plot comparing assembly generated here (red line) with existing yeast reference (black dotted line). Our assembly is slightly smaller (11,287,131 bp versus 12,071,326. Our assembly is lacking the mitochondrial genome (~86 kb) beacuse the initial data does include mitochondrial reads. This is partially responsible for this discrepancy.
With respect to the total sequence length, we can conclude that the size of our genome assembly is very similar to the reference genome. It is noteworthy that the reference genome consists of 17 sequences, while our assembly includes only 16 chromosomes. This is due to the fact that the reference genome also includes the sequence of the mitochondrial DNA, which consists of 85,779 bp. (The above comparison is performed using Quast ( Galaxy version 5.2.0+galaxy1) using Primary assembly generated with scaffolding workflow (WF8) and yeast reference.)
Figure 12: Comparison between contact maps generated using the final Primary assembly from this tutorial (left) and the reference genome (right).
If we compare the contact map of our assembled genome with the reference assembly (Fig. above), we can see that the two are indistinguishable, suggesting that we have generated a chromosome level genome assembly.
Key points
The VGP pipeline allows to generate error-free, near gapless reference-quality genome assemblies
The assembly can be divided in four main stages: genome profile analysis, HiFi long read phased assembly with hifiasm, Bionano hybrid scaffolding and Hi-C hybrid scaffolding
Simão, F. A., R. M. Waterhouse, P. Ioannidis, E. V. Kriventseva, and E. M. Zdobnov, 2015 BUSCO: assessing genome assembly and annotation completeness with single-copy orthologs. Bioinformatics 31: 3210–3212. 10.1093/bioinformatics/btv351
Guan, D., S. A. McCarthy, J. Wood, K. Howe, Y. Wang et al., 2019 Identifying and removing haplotypic duplication in primary genome assemblies. 10.1101/729962
Rhie, A., B. P. Walenz, S. Koren, and A. M. Phillippy, 2020 Merqury: reference-free quality, completeness, and phasing assessment for genome assemblies. Genome Biology 21: 10.1186/s13059-020-02134-9
Glossary
G10K
Genome 10K
GWS
Galaxy Workflow System
Hi-C
all-versus-all chromatin conformation capture
HiFi
high fidelity reads
QV
assembly consensus quality
VGP
Vertebrate Genome Project
alternate assembly
alternate loci not represented in the primary assembly
collection
Galaxy's way to represent multiple datasets as a single interface entity
collections
Galaxy's way to represent multiple datasets as a single interface entity
contigs
contiguous sequences in an assembly
primary assembly
homozygous regions of the genome plus one set of alleles for the heterozygous loci
scaffold
one or more contigs joined by gap sequence
scaffolds
one or more contigs joined by gap sequence
unitig
A uniquely assembleable subset of overlapping fragments. A unitig is an assembly of fragments for which there are no competing internal overlaps. A unitig is either a correctly assembled portion of a contig or a collapsed assembly of several high-fidelity copies of a repeat.
Feedback
Did you use this material as an instructor? Feel free to give us feedback on how it went.
Did you use this material as a learner or student? Click the form below to leave feedback.
Hiltemann, Saskia, Rasche, Helena et al., 2023 Galaxy Training: A Powerful Framework for Teaching! PLOS Computational Biology 10.1371/journal.pcbi.1010752
Batut et al., 2018 Community-Driven Data Analysis Training for Biology Cell Systems 10.1016/j.cels.2018.05.012
@misc{assembly-vgp_workflow_training,
author = "Delphine Lariviere and Alex Ostrovsky and Cristóbal Gallardo and Brandon Pickett and Linelle Abueg and Anton Nekrutenko",
title = "VGP assembly pipeline - short version (Galaxy Training Materials)",
year = "",
month = "",
day = ""
url = "\url{https://training.galaxyproject.org/training-material/topics/assembly/tutorials/vgp_workflow_training/tutorial.html}",
note = "[Online; accessed TODAY]"
}
@article{Hiltemann_2023,
doi = {10.1371/journal.pcbi.1010752},
url = {https://doi.org/10.1371%2Fjournal.pcbi.1010752},
year = 2023,
month = {jan},
publisher = {Public Library of Science ({PLoS})},
volume = {19},
number = {1},
pages = {e1010752},
author = {Saskia Hiltemann and Helena Rasche and Simon Gladman and Hans-Rudolf Hotz and Delphine Larivi{\`{e}}re and Daniel Blankenberg and Pratik D. Jagtap and Thomas Wollmann and Anthony Bretaudeau and Nadia Gou{\'{e}} and Timothy J. Griffin and Coline Royaux and Yvan Le Bras and Subina Mehta and Anna Syme and Frederik Coppens and Bert Droesbeke and Nicola Soranzo and Wendi Bacon and Fotis Psomopoulos and Crist{\'{o}}bal Gallardo-Alba and John Davis and Melanie Christine Föll and Matthias Fahrner and Maria A. Doyle and Beatriz Serrano-Solano and Anne Claire Fouilloux and Peter van Heusden and Wolfgang Maier and Dave Clements and Florian Heyl and Björn Grüning and B{\'{e}}r{\'{e}}nice Batut and},
editor = {Francis Ouellette},
title = {Galaxy Training: A powerful framework for teaching!},
journal = {PLoS Comput Biol} Computational Biology}
}
Congratulations on successfully completing this tutorial!
You can use Ephemeris's shed-tools install command to install the tools used in this tutorial.
Delphine Lariviere
Alex Ostrovsky
Cristóbal Gallardo
Brandon Pickett
Linelle Abueg
Anton NekrutenkoQuestions:
:
Warning: Danger: Make sure you choose corect format!
Open image in new tab
Open image in new tabOpen image in new tab
Open image in new tab
Open image in new tab