Mercurial > repos > earlhaminst > ete
changeset 0:276e3ee68c37 draft
planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/ete commit 230ae552ddeb1bfdef3a09becaa5c6d373529a05-dirty
| author | earlhaminst |
|---|---|
| date | Thu, 15 Dec 2016 12:55:02 -0500 |
| parents | |
| children | a4ba317fc713 |
| files | ete_tree_generator.py ete_tree_generator.xml test-data/out.nhx test-data/species.txt |
| diffstat | 4 files changed, 209 insertions(+), 0 deletions(-) [+] |
line wrap: on
line diff
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/ete_tree_generator.py Thu Dec 15 12:55:02 2016 -0500 @@ -0,0 +1,56 @@ +import optparse + +from ete3 import NCBITaxa + +ncbi = NCBITaxa() + +parser = optparse.OptionParser() +parser.add_option('-s', '--species', dest="input_species_filename", + help='Species list in text format one species in each line') + +parser.add_option('-f', '--format', type='choice', choices=['0', '1', '2', '3', '4', '5', '6', '7', '8', '9', '100'], dest="format", + default='8', help='outpur format for tree') + +parser.add_option('-t', '--treebest', type='choice', choices=['yes', 'no'], dest="treebest", + default='no', help='To be used in TreeBest') + +parser.add_option('-d', '--database', type='choice', choices=['yes', 'no'], dest="database", + default='no', help='Update database') + +options, args = parser.parse_args() + +if options.database == "yes": + try: + ncbi.update_taxonomy_database() + except: + pass + +if options.input_species_filename is None: + raise Exception('-s option must be specified, Species list in text format one species in each line') + +with open(options.input_species_filename) as f: + species_name = [_.strip().replace('_', ' ') for _ in f.readlines()] + +name2taxid = ncbi.get_name_translator(species_name) + +taxid = [name2taxid[_][0] for _ in species_name] + +tree = ncbi.get_topology(taxid) + +if options.treebest == "yes": + inv_map = {str(v[0]): k.replace(" ", "") + "*" for k, v in name2taxid.items()} +else: + inv_map = {str(v[0]): k for k, v in name2taxid.items()} + + +for leaf in tree: + leaf.name = inv_map[leaf.name] + +newickTree = tree.write(format=int(options.format)) + +if options.treebest == "yes": + newickTree = newickTree.rstrip(';') + newickTree = newickTree + "root;" + +with open('newickTree.nhx', 'w') as newickFile: + newickFile.write(newickTree)
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/ete_tree_generator.xml Thu Dec 15 12:55:02 2016 -0500 @@ -0,0 +1,90 @@ +<tool id="species_tree_generator" name="Species tree generator" version="3.0.0b35"> + <description>from a list of species using the ETE Toolkit</description> + <requirements> + <requirement type="package" version="3.0.0b35">ete3</requirement> + </requirements> + <stdio> + <!-- Anything other than zero is an error --> + <exit_code range="1:" /> + </stdio> + <command> + <![CDATA[ +python $__tool_directory__/ete_tree_generator.py +-s '$speciesFile' +#if $output_format.treebest == 'yes' + -f 8 +#else + -f ${output_format.format_selector} +#end if +-t $output_format.treebest +-d $database + ]]> + </command> + <inputs> + <param name="speciesFile" type="data" format="txt" label="Species file" help="List with one species per line" /> + <conditional name="output_format"> + <param name="treebest" type="select" label="Use in TreeBest" help="Select yes if specie tree to be used in TreeBest"> + <option value="yes">Yes</option> + <option value="no" selected="true">No</option> + </param> + <when value="yes" /> + <when value="no"> + <param name="format_selector" type="select" label="Output format" help="See table below"> + <option value="0" selected="true">Flexible with support values (0)</option> + <option value="1">Flexible with internal node names (1)</option> + <option value="2">All branches + leaf names + internal supports (2)</option> + <option value="3">All branches + all names (3)</option> + <option value="4">Leaf branches + leaf names (4)</option> + <option value="5">Internal and leaf branches + leaf names (5)</option> + <option value="6">Internal branches + leaf names (6)</option> + <option value="7">Leaf branches + all names (7)</option> + <option value="8">All names (8)</option> + <option value="9">Leaf names (9)</option> + <option value="100">Topology only (100)</option> + </param> + </when> + </conditional> + <param name="database" type="select" label="Update database" help="Update database from NCBI (slower)"> + <option value="no" selected="true">No</option> + <option value="yes">Yes</option> + </param> + </inputs> + <outputs> + <data format="nhx" name="outputFile" label="${tool.name} on ${on_string}" from_work_dir="newickTree.nhx"/> + </outputs> + <tests> + <test> + <param name="speciesFile" ftype="txt" value="species.txt" /> + <param name="treebest" value="yes" /> + <output name="outputFile" file="out.nhx" /> + </test> + </tests> + <help> + <![CDATA[ +Generate a species tree from a list of species using the `ETE Toolkit`_. + +.. _ETE Toolkit: http://etetoolkit.org/ + +**Output format:** + +======= ============================================= ======================================================================================== +FORMAT DESCRIPTION SAMPLE +------- --------------------------------------------- ---------------------------------------------------------------------------------------- +0 flexible with support values ((D:0.723274,F:0.567784)1.000000:0.067192,(B:0.279326,H:0.756049)1.000000:0.807788); +1 flexible with internal node names ((D:0.723274,F:0.567784)E:0.067192,(B:0.279326,H:0.756049)B:0.807788); +2 all branches + leaf names + internal supports ((D:0.723274,F:0.567784)1.000000:0.067192,(B:0.279326,H:0.756049)1.000000:0.807788); +3 all branches + all names ((D:0.723274,F:0.567784)E:0.067192,(B:0.279326,H:0.756049)B:0.807788); +4 leaf branches + leaf names ((D:0.723274,F:0.567784),(B:0.279326,H:0.756049)); +5 internal and leaf branches + leaf names ((D:0.723274,F:0.567784):0.067192,(B:0.279326,H:0.756049):0.807788); +6 internal branches + leaf names ((D,F):0.067192,(B,H):0.807788); +7 leaf branches + all names ((D:0.723274,F:0.567784)E,(B:0.279326,H:0.756049)B); +8 all names ((D,F)E,(B,H)B); +9 leaf names ((D,F),(B,H)); +100 topology only ((,),(,)); +======= ============================================= ======================================================================================== + ]]> + </help> + <citations> + <citation type="doi">10.1093/molbev/msw046</citation> + </citations> +</tool>
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/test-data/out.nhx Thu Dec 15 12:55:02 2016 -0500 @@ -0,0 +1,1 @@ +((((((((((((((((Homosapiens*,Pantroglodytes*,Gorillagorillagorilla*)207598,Pongoabelii*)9604,Nomascusleucogenys*)314295,(Chlorocebussabaeus*,Macacamulatta*,Papioanubis*)9528)9526,Callithrixjacchus*)314293,Tarsiussyrichta*)376913,(Microcebusmurinus*,Otolemurgarnettii*)376911)9443,((Oryctolaguscuniculus*,Ochotonaprinceps*)9975,(((Rattusnorvegicus*,Musmusculus*)39107,Dipodomysordii*,Ictidomystridecemlineatus*)33553,Caviaporcellus*)9989)314147,Tupaiabelangeri*)314146,(((Ailuropodamelanoleuca*,Canislupusfamiliaris*,Mustelaputoriusfuro*)379584,Feliscatus*)33554,(Erinaceuseuropaeus*,Sorexaraneus*)9362,(Myotislucifugus*,Pteropusvampyrus*)9397,((Ovisaries*,Bostaurus*)9895,Vicugnapacos*,Susscrofa*,Tursiopstruncatus*)91561,Equuscaballus*)314145)1437010,(Dasypusnovemcinctus*,Choloepushoffmanni*)9348,(Echinopstelfairi*,Loxodontaafricana*,Procaviacapensis*)311790)9347,(Sarcophilusharrisii*,Macropuseugenii*,Monodelphisdomestica*)9263)32525,Ornithorhynchusanatinus*)40674,((Pelodiscussinensis*,(((Meleagrisgallopavo*,Gallusgallus*)9005,Anasplatyrhynchos*)1549675,(Ficedulaalbicollis*,Taeniopygiaguttata*)9126)8825)1329799,Anoliscarolinensis*)32561)32524,Xenopustropicalis*)32523,Latimeriachalumnae*)8287,(((Daniorerio*,Astyanaxmexicanus*)186626,((((Tetraodonnigroviridis*,Takifugurubripes*)31031,Gasterosteusaculeatus*)1489922,((Poeciliaformosa*,Xiphophorusmaculatus*)586240,Oreochromisniloticus*)1489908)1489872,Gadusmorhua*)123368)186625,Lepisosteusoculatus*)41665)root; \ No newline at end of file
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/test-data/species.txt Thu Dec 15 12:55:02 2016 -0500 @@ -0,0 +1,62 @@ +Latimeria chalumnae +Xenopus tropicalis +Ornithorhynchus anatinus +Macropus eugenii +Sarcophilus harrisii +Monodelphis domestica +Echinops telfairi +Procavia capensis +Loxodonta africana +Dasypus novemcinctus +Choloepus hoffmanni +Ochotona princeps +Oryctolagus cuniculus +Cavia porcellus +Ictidomys tridecemlineatus +Dipodomys ordii +Rattus norvegicus +Mus musculus +Tupaia belangeri +Otolemur garnettii +Microcebus murinus +Tarsius syrichta +Callithrix jacchus +Chlorocebus sabaeus +Papio anubis +Macaca mulatta +Nomascus leucogenys +Pongo abelii +Gorilla gorilla gorilla +Pan troglodytes +Homo sapiens +Sorex araneus +Erinaceus europaeus +Pteropus vampyrus +Myotis lucifugus +Felis catus +Canis lupus familiaris +Mustela putorius furo +Ailuropoda melanoleuca +Equus caballus +Vicugna pacos +Sus scrofa +Tursiops truncatus +Ovis aries +Bos taurus +Anolis carolinensis +Pelodiscus sinensis +Taeniopygia guttata +Ficedula albicollis +Anas platyrhynchos +Meleagris gallopavo +Gallus gallus +Lepisosteus oculatus +Danio rerio +Astyanax mexicanus +Gadus morhua +Oreochromis niloticus +Poecilia formosa +Xiphophorus maculatus +Gasterosteus aculeatus +Tetraodon nigroviridis +Takifugu rubripes