The number of genome assemblies generated and stored in sequence databases is growing exponentially every year. With the availability of this growing amount of data, meta-genomics studies become more and more popular. By using this bulk of genomes for comparing them to thousands of other genomes new structural patterns and evolutionary insights can be obtained. However, the first step in any meta-genomics study is the retrieval of the genomes, proteomes, coding sequences or annotation files that shall be compared and investigated.
For this purpose, the meta.retrieval() function allows users to perform straightforward meta-genome retrieval in R.
The getKingdoms() function stores a list of all available kingdoms of life. Using the argument db users can specify from which database kingdom information shall be retrieved.
Example RefSeq:
getKingdoms(db = "refseq")[1] "archaea" "bacteria" "fungi" "invertebrate"
[5] "plant" "protozoa" "vertebrate_mammalian" "vertebrate_other"
[9] "viral"Example Genbank:
getKingdoms(db = "genbank")[1] "archaea" "bacteria" "fungi"
[4] "invertebrate" "plant" "protozoa"
[7] "vertebrate_mammalian" "vertebrate_other"In these examples the difference betwenn db = "refseq" and db = "genbank" is that db = "genbank" does not store viral information.
These kingdoms can be specified in meta.retrieval().
The meta.retrieval() function aims to simplify the genome retrieval process for subsequent meta-genomics studies.
Usually this step is performed with shell scripts. However, since many meta-genomics packages exist for the R programming language, I implemented this functionality for easy integration into existing workflows.
For example, the pipeline logic of the magrittr package can be used with meta.retrieval().
# download all vertebrate genomes, then apply ...
meta.retrieval(kingdom = "vertebrate_mammalian", db = "refseq", type = "genome") %>% ...Here ... denotes any subsequent meta-genomics analysis. Hence, meta.retrieval() enables the pipelining methodology for meta-genomics.
The meta.retrieval() function can retrieve genomes, proteomes, and CDS files.
Download all mammalian vertebrate genomes from RefSeq.
# download all vertebrate genomes
meta.retrieval(kingdom = "vertebrate_mammalian", db = "refseq", type = "genome")All geneomes are stored in the folder named according to the kingdom. In this case vertebrate_mammalian. Alternatively, users can specify the out.folder argument to define a custom output folder path.
Example Bacteria
# download all bacteria genomes
meta.retrieval(kingdom = "bacteria", db = "refseq", type = "genome")Example Viruses
# download all virus genomes
meta.retrieval(kingdom = "viral", db = "refseq", type = "genome")Example Archaea
# download all archaea genomes
meta.retrieval(kingdom = "archaea", db = "refseq", type = "genome")Example Fungi
# download all fungi genomes
meta.retrieval(kingdom = "fungi", db = "refseq", type = "genome")Example Plants
# download all plant genomes
meta.retrieval(kingdom = "plant", db = "refseq", type = "genome")Example Invertebrates
# download all invertebrate genomes
meta.retrieval(kingdom = "invertebrate", db = "refseq", type = "genome")Example Protozoa
# download all invertebrate genomes
meta.retrieval(kingdom = "protozoa", db = "refseq", type = "genome")Alternatively, download all mammalian vertebrate genomes from Genbank, e.g.
# download all vertebrate genomes
meta.retrieval(kingdom = "vertebrate_mammalian", db = "genbank", type = "genome")Gammaproteobacteria genomes from NCBI RefSeq:Please note that the following functionality is so far only provided by the developer version of biomartr.
In case users do not wish to retrieve genomes from an entire kingdom, but rather from a subgoup (e.g. from species belonging to the Gammaproteobacteria class, a subgroup of the bacteria kingdom), they can use the following workflow:
First, users can again consult the getKingdoms() function to retrieve kingdom information.
getKingdoms(db = "refseq")[1] "archaea" "bacteria" "fungi" "invertebrate"
[5] "plant" "protozoa" "vertebrate_mammalian" "vertebrate_other"
[9] "viral"In this example, we will choose the bacteria kingdom. Now, the getGroups() function allows users to obtain available subgroups of the bacteria kingdom.
getGroups(db = "refseq", kingdom = "bacteria") [1] "Acidithiobacillia" "Acidobacteriia"
[3] "Actinobacteria" "Alphaproteobacteria"
[5] "Aquificae" "Armatimonadetes"
[7] "Bacteroidetes/Chlorobi group" "Balneolia"
[9] "Betaproteobacteria" "Blastocatellia"
[11] "Candidatus Kryptonia" "Chlamydiae"
[13] "Chloroflexi" "Cyanobacteria/Melainabacteria group"
[15] "Deinococcus-Thermus" "delta/epsilon subdivisions"
[17] "Endomicrobia" "Fibrobacteres"
[19] "Firmicutes" "Fusobacteriia"
[21] "Gammaproteobacteria" "Gemmatimonadetes"
[23] "Kiritimatiellaeota" "Nitrospira"
[25] "Planctomycetes" "Spirochaetia"
[27] "Synergistia" "Tenericutes"
[29] "Thermodesulfobacteria" "Thermotogae"
[31] "unclassified Acidobacteria" "unclassified Bacteria (miscellaneous)"
[33] "unclassified Proteobacteria" "Verrucomicrobia"
[35] "Zetaproteobacteria" Please note, that the kingdom argument specified in getGroups() needs to match with an available kingdom retrieved with getKingdoms(). It is also important that in both cases: getKingdoms() and getGroups() the same database should be specified.
Now we choose the group Gammaproteobacteria and specify the group argument in the meta.retrieval() function.
meta.retrieval(kingdom = "bacteria", group = "Gammaproteobacteria", db = "refseq", type = "genome")Using this command, all bacterial (kingdom = "bacteria") genomes (type = "genome") that belong to the group Gammaproteobacteria (group = "Gammaproteobacteria") will be retrieved from NCBI RefSeq (db = "refseq").
Alternatively, Gammaproteobacteria genomes can be retrieved from NCBI Genbank by exchanging db = "refseq" to db = "genbank". If users wish to download proteome, CDS, or GFF files instead of genomes, they can specify the argument: type = "proteome", type = "CDS", or type = "gff".
Adenoviridae genomes from NCBI RefSeq:Retrieve groups for viruses.
getGroups(db = "refseq", kingdom = "viral") [1] "Adenoviridae" "Alloherpesviridae"
[3] "Alphaflexiviridae" "Alphatetraviridae"
[5] "Alvernaviridae" "Amalgaviridae"
[7] "Ampullaviridae" "Anelloviridae"
[9] "Apple fruit crinkle viroid" "Apple hammerhead viroid-like circular RNA"
[11] "Apscaviroid" "Arenaviridae"
[13] "Arteriviridae" "Ascoviridae"
[15] "Asfarviridae" "Astroviridae"
[17] "Avsunviroid" "Baculoviridae"
[19] "Barnaviridae" "Benyviridae"
[21] "Betaflexiviridae" "Bicaudaviridae"
[23] "Birnaviridae" "Bornaviridae"
[25] "Bromoviridae" "Bunyaviridae"
[27] "Caliciviridae" "Carmotetraviridae"
[29] "Caulimoviridae" "Cherry leaf scorch small circular viroid-like RNA 1"
[31] "Cherry small circular viroid-like RNA" "Chrysoviridae"
[33] "Circoviridae" "Closteroviridae"
[35] "Cocadviroid" "Coleviroid"
[37] "Coronaviridae" "Corticoviridae"
[39] "Cystoviridae" "Dicistroviridae"
[41] "Elaviroid" "Endornaviridae"
[43] "Filoviridae" "Flaviviridae"
[45] "Fusarividae" "Fuselloviridae"
[47] "Gammaflexiviridae" "Geminiviridae"
[49] "Genomoviridae" "Globuloviridae"
[51] "Grapevine latent viroid" "Guttaviridae"
[53] "Hepadnaviridae" "Hepeviridae"
[55] "Herpesviridae" "Hostuviroid"
[57] "Hypoviridae" "Hytrosaviridae"
[59] "Iflaviridae" "Inoviridae"
[61] "Iridoviridae" "Lavidaviridae"
[63] "Leviviridae" "Lipothrixviridae"
[65] "Luteoviridae" "Malacoherpesviridae"
[67] "Marnaviridae" "Marseilleviridae"
[69] "Megabirnaviridae" "Mesoniviridae"
[71] "Microviridae" "Mimiviridae"
[73] "Mulberry small circular viroid-like RNA 1" "Mymonaviridae"
[75] "Myoviridae" "Nanoviridae"
[77] "Narnaviridae" "Nimaviridae"
[79] "Nodaviridae" "Nudiviridae"
[81] "Nyamiviridae" "Ophioviridae"
[83] "Orthomyxoviridae" "Other"
[85] "Papillomaviridae" "Paramyxoviridae"
[87] "Partitiviridae" "Parvoviridae"
[89] "Pelamoviroid" "Permutotetraviridae"
[91] "Persimmon viroid" "Phycodnaviridae"
[93] "Picobirnaviridae" "Picornaviridae"
[95] "Plasmaviridae" "Pneumoviridae"
[97] "Podoviridae" "Polydnaviridae"
[99] "Polyomaviridae" "Pospiviroid"
[101] "Potyviridae" "Poxviridae"
[103] "Quadriviridae" "Reoviridae"
[105] "Retroviridae" "Rhabdoviridae"
[107] "Roniviridae" "Rubber viroid India/2009"
[109] "Rudiviridae" "Secoviridae"
[111] "Siphoviridae" "Sphaerolipoviridae"
[113] "Sunviridae" "Tectiviridae"
[115] "Togaviridae" "Tombusviridae"
[117] "Totiviridae" "Turriviridae"
[119] "Tymoviridae" "unclassified"
[121] "unclassified Pospiviroidae" "Virgaviridae"Now we can choose Adenoviridae as group argument for the meta.retrieval() function.
meta.retrieval(kingdom = "viral", group = "Adenoviridae", db = "refseq", type = "genome")Again, by exchanging type = "genome" by either type = "proteome", type = "CDS", or type = "gff", users can retrieve proteome, CDS, or GFF files instead of genomes.
Although much effort is invested to increase the genome assembly quality when new genomes are published or new versions are released, the influence of genome assembly quality on downstream analyses cannot be neglected. A rule of thumb is, that the larger the genome the more prone it is to genome assembly errors and therefore, a reduction of assembly quality.
In Veeckman et al., 2016 the authors conclude:
As yet, no uniform metrics or standards are in place to estimate the completeness of a genome assembly or the annotated gene space, despite their importance for downstream analyses
In most metagenomics studies, however, the influence or bias of genome assembly quality on the outcome of the analysis (e.g. comparative genomics, annotation, etc.) is neglected. To better grasp the genome assembly quality, the NCBI databases store genome assembly statistics of some species for which genome assemblies are available. An example assembly statistics report can be found at: ftp://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/000/001/405/GCF_000001405.36_GRCh38.p10/GCF_000001405.36_GRCh38.p10_assembly_stats.txt.
The biomartr package allows users to retrieve these genome assembly stats file in an automated way.
# show all elements of the data.frame
options(tibble.print_max = Inf)
# retrieve genome assembly stats for all mammal genome assemblies
# and store these stats in a data.frame
mammals.gc <- meta.retrieval(kingdom = "vertebrate_mammalian",
db = "refseq",
type = "assemblystats",
combine = TRUE)
mammals.gc species total_length spanned_gaps unspanned_gaps region_count scaffold_count
<chr> <int> <int> <int> <int> <int>
1 Ornithorhynchus anatinus 1995607322 243698 137 0 200283
2 Sarcophilus harrisii NA 201317 0 0 35974
3 Dasypus novemcinctus NA 268413 0 0 46559
4 Erinaceus europaeus NA 219764 0 0 5803
5 Echinops telfairi NA 269444 0 0 8402
6 Pteropus alecto 1985975446 104566 0 0 65598
7 Rousettus aegyptiacus 1910250568 559 0 0 NA
8 Callithrix jacchus NA 184972 2242 0 16399
9 Cebus capucinus imitator NA 133441 0 0 7156
10 Cercocebus atys NA 65319 0 0 11433
# ... with 89 more rows, and 9 more variables: scaffold_N50 <int>, scaffold_L50 <int>,
# scaffold_N75 <int>, scaffold_N90 <int>, contig_count <int>, contig_N50 <int>, total_gap_length <int>,
# molecule_count <int>, top_level_count <int>Analogous, for each kingdom available at getKingdoms() this information can be retrieved.
NCBI Genbank stores metagenome projects in addition to species specific genome, proteome or CDS sequences. To retrieve these metagenomes users can perform the following combination of commands:
First, users can list the project names of available metagenomes by typing
# list available metagenomes at NCBI Genbank
listMetaGenomes()[1] "metagenome" "human gut metagenome" "epibiont metagenome"
[4] "marine metagenome" "soil metagenome" "mine drainage metagenome"
[7] "mouse gut metagenome" "marine sediment metagenome" "termite gut metagenome"
[10] "hot springs metagenome" "human lung metagenome" "fossil metagenome"
[13] "freshwater metagenome" "saltern metagenome" "stromatolite metagenome"
[16] "coral metagenome" "mosquito metagenome" "fish metagenome"
[19] "bovine gut metagenome" "chicken gut metagenome" "wastewater metagenome"
[22] "microbial mat metagenome" "freshwater sediment metagenome" "human metagenome"
[25] "hydrothermal vent metagenome" "compost metagenome" "wallaby gut metagenome"
[28] "groundwater metagenome" "gut metagenome" "sediment metagenome"
[31] "ant fungus garden metagenome" "food metagenome" "hypersaline lake metagenome"
[34] "hydrocarbon metagenome" "activated sludge metagenome" "viral metagenome"
[37] "bioreactor metagenome" "wasp metagenome" "permafrost metagenome"
[40] "sponge metagenome" "aquatic metagenome" "insect gut metagenome"
[43] "activated carbon metagenome" "anaerobic digester metagenome" "rock metagenome"
[46] "terrestrial metagenome" "rock porewater metagenome" "seawater metagenome"
[49] "scorpion gut metagenome" "soda lake metagenome" "glacier metagenome"Internally the listMetaGenomes() function downloads the assembly_summary.txt file from ftp://ftp.ncbi.nlm.nih.gov/genomes/genbank/metagenomes/ to retrieve available metagenome information. This procedure might take a few seconds during the first run of listMetaGenomes(). Subsequently, the assembly_summary.txt file will be stored in the tempdir() directory to achieve a much faster access of this information during following uses of listMetaGenomes().
In case users wish to retrieve detailed information about available metagenome projects they can specify the details = TRUE argument.
# show all elements of the data.frame
options(tibble.print_max = Inf)
# detailed information on available metagenomes at NCBI Genbank
listMetaGenomes(details = TRUE)# A tibble: 857 x 21
assembly_accession bioproject biosample wgs_master refseq_category taxid species_taxid
<chr> <chr> <chr> <chr> <chr> <int> <int>
1 GCA_000206185.1 PRJNA32359 SAMN02954317 AAGA00000000.1 na 256318 256318
2 GCA_000206205.1 PRJNA32355 SAMN02954315 AAFZ00000000.1 na 256318 256318
3 GCA_000206225.1 PRJNA32357 SAMN02954316 AAFY00000000.1 na 256318 256318
4 GCA_000208265.2 PRJNA17779 SAMN02954240 AASZ00000000.1 na 256318 256318
5 GCA_000208285.1 PRJNA17657 SAMN02954268 AATO00000000.1 na 256318 256318
6 GCA_000208305.1 PRJNA17659 SAMN02954269 AATN00000000.1 na 256318 256318
7 GCA_000208325.1 PRJNA16729 SAMN02954263 AAQL00000000.1 na 256318 256318
8 GCA_000208345.1 PRJNA16729 SAMN02954262 AAQK00000000.1 na 256318 256318
9 GCA_000208365.1 PRJNA13699 SAMN02954283 AAFX00000000.1 na 256318 256318
10 GCA_900010595.1 PRJEB11544 SAMEA3639840 CZPY00000000.1 na 256318 256318
# ... with 847 more rows, and 14 more variables: organism_name <chr>, infraspecific_name <chr>,
# isolate <chr>, version_status <chr>, assembly_level <chr>, release_type <chr>, genome_rep <chr>,
# seq_rel_date <date>, asm_name <chr>, submitter <chr>, gbrs_paired_asm <chr>, paired_asm_comp <chr>,
# ftp_path <chr>, excluded_from_refseq <chr>Finally, users can retrieve available metagenomes using getMetaGenomes(). The name argument receives the metagenome project name retrieved with listMetaGenomes(). The path argument specifies the folder path in which corresponding genomes shall be stored.
# retrieve all genomes belonging to the human gut metagenome project
getMetaGenomes(name = "human gut metagenome", path = file.path("_ncbi_downloads","human_gut"))1] "The metagenome of 'human gut metagenome' has been downloaded to '_ncbi_downloads/human_gut'."
[1] "_ncbi_downloads/human_gut/GCA_000205525.2_ASM20552v2_genomic.fna.gz"
[2] "_ncbi_downloads/human_gut/GCA_000205765.1_ASM20576v1_genomic.fna.gz"
[3] "_ncbi_downloads/human_gut/GCA_000205785.1_ASM20578v1_genomic.fna.gz"
[4] "_ncbi_downloads/human_gut/GCA_000207925.1_ASM20792v1_genomic.fna.gz"
[5] "_ncbi_downloads/human_gut/GCA_000207945.1_ASM20794v1_genomic.fna.gz"
[6] "_ncbi_downloads/human_gut/GCA_000207965.1_ASM20796v1_genomic.fna.gz"
[7] "_ncbi_downloads/human_gut/GCA_000207985.1_ASM20798v1_genomic.fna.gz"
[8] "_ncbi_downloads/human_gut/GCA_000208005.1_ASM20800v1_genomic.fna.gz"
[9] "_ncbi_downloads/human_gut/GCA_000208025.1_ASM20802v1_genomic.fna.gz"
[10] "_ncbi_downloads/human_gut/GCA_000208045.1_ASM20804v1_genomic.fna.gz"
[11] "_ncbi_downloads/human_gut/GCA_000208065.1_ASM20806v1_genomic.fna.gz"
[12] "_ncbi_downloads/human_gut/GCA_000208085.1_ASM20808v1_genomic.fna.gz"
[13] "_ncbi_downloads/human_gut/GCA_000208105.1_ASM20810v1_genomic.fna.gz"
[14] "_ncbi_downloads/human_gut/GCA_000208125.1_ASM20812v1_genomic.fna.gz"
[15] "_ncbi_downloads/human_gut/GCA_000208145.1_ASM20814v1_genomic.fna.gz"
[16] "_ncbi_downloads/human_gut/GCA_000208165.1_ASM20816v1_genomic.fna.gz"
...Internally, getMetaGenomes() creates a folder specified in the path argument. Genomes associated with the metagenomes project specified in the name argument will then be downloaded and stored in this folder. As return value getMetaGenomes() returns the file paths to the genome files which can then be used as input to the read*() functions.
Alternatively or subsequent to the metagenome retrieval, users can retrieve annotation files of genomes belonging to a metagenome project selected with listMetaGenomes() by using the getMetaGenomeAnnotations() function.
# retrieve all genomes belonging to the human gut metagenome project
getMetaGenomeAnnotations(name = "human gut metagenome", path = file.path("_ncbi_downloads","human_gut","annotations"))[1] "The annotations of metagenome 'human gut metagenome' have been downloaded and stored at '_ncbi_downloads/human_gut/annotations'."
[1] "_ncbi_downloads/human_gut/annotations/GCA_000205525.2_ASM20552v2_genomic.gff.gz"
[2] "_ncbi_downloads/human_gut/annotations/GCA_000205765.1_ASM20576v1_genomic.gff.gz"
[3] "_ncbi_downloads/human_gut/annotations/GCA_000205785.1_ASM20578v1_genomic.gff.gz"
[4] "_ncbi_downloads/human_gut/annotations/GCA_000207925.1_ASM20792v1_genomic.gff.gz"
[5] "_ncbi_downloads/human_gut/annotations/GCA_000207945.1_ASM20794v1_genomic.gff.gz"
[6] "_ncbi_downloads/human_gut/annotations/GCA_000207965.1_ASM20796v1_genomic.gff.gz"
[7] "_ncbi_downloads/human_gut/annotations/GCA_000207985.1_ASM20798v1_genomic.gff.gz"
[8] "_ncbi_downloads/human_gut/annotations/GCA_000208005.1_ASM20800v1_genomic.gff.gz"
[9] "_ncbi_downloads/human_gut/annotations/GCA_000208025.1_ASM20802v1_genomic.gff.gz"
[10] "_ncbi_downloads/human_gut/annotations/GCA_000208045.1_ASM20804v1_genomic.gff.gz"
[11] "_ncbi_downloads/human_gut/annotations/GCA_000208065.1_ASM20806v1_genomic.gff.gz"
[12] "_ncbi_downloads/human_gut/annotations/GCA_000208085.1_ASM20808v1_genomic.gff.gz"
[13] "_ncbi_downloads/human_gut/annotations/GCA_000208105.1_ASM20810v1_genomic.gff.gz"
[13] "_ncbi_downloads/human_gut/annotations/GCA_000208105.1_ASM20810v1_genomic.gff.gz"
[14] "_ncbi_downloads/human_gut/annotations/GCA_000208125.1_ASM20812v1_genomic.gff.gz"
[15] "_ncbi_downloads/human_gut/annotations/GCA_000208145.1_ASM20814v1_genomic.gff.gz"
[16] "_ncbi_downloads/human_gut/annotations/GCA_000208165.1_ASM20816v1_genomic.gff.gz"
...The file paths of the downloaded *.gff are retured by getMetaGenomeAnnotations() and can be used as input for the read.gff() function in the seqreadr package.
Download all mammalian vertebrate proteomes.
Example RefSeq:
# download all vertebrate genomes
meta.retrieval(kingdom = "vertebrate_mammalian", db = "refseq", type = "proteome")Example Genbank:
# download all vertebrate genomes
meta.retrieval(kingdom = "vertebrate_mammalian", db = "genbank", type = "proteome")Download all mammalian vertebrate CDS from RefSeq (Genbank does not store CDS data).
Example RefSeq:
# download all vertebrate genomes
meta.retrieval(kingdom = "vertebrate_mammalian", db = "refseq", type = "CDS")Example Genbank:
# download all vertebrate genomes
meta.retrieval(kingdom = "vertebrate_mammalian", db = "genbank", type = "CDS")Download all mammalian vertebrate gff files.
Example RefSeq:
# download all vertebrate gff files
meta.retrieval(kingdom = "vertebrate_mammalian", db = "refseq", type = "gff")Example Genbank:
# download all vertebrate gff files
meta.retrieval(kingdom = "vertebrate_mammalian", db = "genbank", type = "gff")Users can obtain alternative kingdoms using getKingdoms().
If users wish to download the all genomes, proteome, CDS, or gff files for all species available in RefSeq or Genbank, they can use the meta.retrieval.all() function for this purpose.
Example RefSeq:
# download all geneomes stored in RefSeq
meta.retrieval.all(db = "refseq", type = "genome")Example Genbank:
# download all geneomes stored in Genbank
meta.retrieval.all(db = "genbank", type = "genome")Example RefSeq:
# download all proteome stored in RefSeq
meta.retrieval.all(db = "refseq", type = "proteome")Example Genbank:
# download all proteome stored in Genbank
meta.retrieval.all(db = "genbank", type = "proteome")