+[c @ s d Z d d l Z d d l Z d d l Z d d l Z d d l Z d d l Z d d l m Z d d l m Z d d l
m Z d d l m
Z
d Z d Z e d k r e n d S(
s
.. module:: run_BUSCO
:synopsis: BUSCO - Benchmarking Universal Single-Copy Orthologs.
.. versionadded:: 3.0.0
.. versionchanged:: 3.0.1
This is the BUSCO main script.
To get help, ``python run_BUSCO.py -h``. See also the user guide.
And visit our website ``_
Copyright (c) 2016-2017, Evgeny Zdobnov (ez@ezlab.org)
Licensed under the MIT license. See LICENSE.md file.
iN( t BuscoConfig( t RawTextHelpFormatter( t isdir( t is_non_empty_filec C s0 t j d d t j d d d t d t } | j d } | j d d d
d d t d
d d d | j d d d
d d t d
d d d | j d d d
d d t d
d d d | j d d d
d d t d
d d t d d t j d | j d d! d
d" d t d
d# d d$ | j d% d& d
d' d t d
d( d d) | j d* d+ d, d- d t d
d. d d/ | j d0 d1 d, d- d t d
d2 d d3 | j d4 d5 d t d
d6 d
d7 d d8 | j d9 d t d
d: d d; | j d< d= d
d> d t d
d? d d@ t j d? | j dA d
dB d
dC d t d t
d dD t t j dB | j dE d, d- d t d
dF d dG | j dH dI d
dJ d t d dK d, d- | j dL dM d
dN d t d dO d, d- | j dP d
dQ d t d dR d, d- | j dS dT d, dU d dV dU dW t j | j dX dY d, d d dZ t | j
S([ s
This function parses the arguments provided by the user
:return: a dictionary having a key for each arguments
:rtype: dict
t descriptions Welcome to BUSCO %s: the Benchmarking Universal Single-Copy Ortholog assessment tool.
For more detailed usage information, please review the README file provided with this distribution and the BUSCO user guide.t usagesZ python BUSCO.py -i [SEQUENCE_FILE] -l [LINEAGE] -o [OUTPUT_NAME] -m [MODE] [OTHER OPTIONS]t formatter_classt add_helps optional argumentss -is --int destt int requiredt metavars
FASTA FILEt helps Input sequence file in FASTA format. Can be an assembled genome or transcriptome (DNA), or protein sequences from an annotated gene set.s -cs --cput cput Ns7 Specify the number (N=integer) of threads/cores to use.s -os --outt outt OUTPUTs Give your analysis run a recognisable short name. Output folders and files will be labelled with this name. WARNING: do not provide a paths -es --evaluet evaluet typesR E-value cutoff for BLAST searches. Allowed formats, 0.001 or 1e-03 (Default: %.0e)s -ms --modet modet MODEs Specify which BUSCO analysis mode to run.
There are three valid modes:
- geno or genome, for genome assemblies (DNA)
- tran or transcriptome, for transcriptome assemblies (DNA)
- prot or proteins, for annotated gene sets (protein)s -ls --lineage_patht lineage_patht LINEAGEsk Specify location of the BUSCO lineage data to be used.
Visit http://busco.ezlab.org for available lineages.s -fs --forcet actiont
store_truet forcesg Force rewriting of existing files. Must be used when output files with the provided name already exist.s -rs --restartt restarts> Restart an uncompleted run. Not available for the protein modes -sps --speciest speciest SPECIESsl Name of existing Augustus species gene finding parameters. See Augustus documentation for available options.s --augustus_parameterst augustus_parameterss Additional parameters for the fine-tuning of Augustus run. For the species, do not use this option.
Use single quotes as follow: '--param1=1 --param2=2', see Augustus documentation for available options.s -ts
--tmp_patht PATHt tmp_paths, Where to store temporary files (Default: %s)s --limitt limitt REGION_LIMITsU How many candidate regions (contig or transcript) to consider per BUSCO (default: %s)s --longt longs Optimization mode Augustus self-training (Default: Off) adds considerably to the run time, but can improve results for some non-model organismss -qs --quiett quiets+ Disable the info logs, displays only errorss -zs --tarzipt tarzips> Tarzip the output folders likely to contain thousands of filess --blast_single_coret blast_single_cores Force tblastn to run on a single core and ignore the --cpu argument for this step only. Useful if inconsistencies when using multiple threads are noticeds -vs --versiont versions Show this version and exits BUSCO %ss -hs --helps Show this help message and exit( t argparset ArgumentParserR t VERSIONR t Falset add_argument_groupt add_argumentt floatt DEFAULT_ARGS_VALUESt intt strt varst
parse_args( t parsert optional( ( sD /home/psgendb/BIRCHDEV/install/quast-5.2.0/quast_libs/busco/busco.pyt _parse_args# st
""! c C sF t j } t j j d rT t j t j j d t j rT t j j d } n" d t j j t j j t } t
| d i | d 6| d 6} t j j | j d d d | } t | s t j
| n t j j | d | } d
d l m } | | j _ d
d l m } d
d
l m } d
d l m }
| j j | j } t | rw| j d | d | Sy y | j d t
j t j d f | j d | | | } | j | j s| j d t t j | n! | j d t t j | | j d | j! Wn& |
k
rU}
| j" |
t# n XWn t# k
r| j" d | j" d t
j$ t# n t% k
r| j" d | j" d | j" d t
j$ t# nm t& k
rAt' j( \ } } } | j) d t* j+ | | | | j" d | j" d t
j$ t# n X| S( s
This function runs a BUSCO analysis according to the provided parameters.
See the help for more details:
``python run_BUSCO.py -h``
:raises SystemExit: if any errors occur
t BUSCO_CONFIG_FILEs %s//config.ini.defaultt argsR R t buscot out_paths run_%ss short_summary_%s.txti( t
pipebricks( t GenomeAnalysis( t
BuscoAnalysis( t
ToolExceptions Using existing BUSCO files for s ...sQ ****************** Start a BUSCO %s analysis, current time: %s ******************s %m/%d/%Y %H:%M:%Ss Configuration loaded from %ss3 BUSCO analysis done. Total running time: %s secondssC BUSCO analysis done with WARNING(s). Total running time: %s secondss Results written in %s
s BUSCO analysis failed !sa Check the logs, read the user guide, if you still need technical support, then please contact %s
s% A signal was sent to kill the processs! Unhandled exception occurred: %s
(, t timet ost environt gett accesst R_OKt patht dirnamet realpatht __file__R t joinR t makedirst quast_libs.buscoR: t
PipeLoggert run_dirpatht quast_libs.busco.GenomeAnalysisR; t quast_libs.busco.BuscoAnalysisR< t# quast_libs.busco.pipebricks.ToolsetR= t _loggert
reload_logR t infoR) t strftimet run_analysist has_warningR0 t mainoutt errort
SystemExitt CONTACTt KeyboardInterruptt
BaseExceptiont syst exc_infot criticalt tracebackt format_exception( t in_fpatht out_fnamet
start_timet config_filet configt assembly_dirpatht summary_pathR: R; R< R= t loggert analysist et exc_typet exc_valuet
exc_traceback( ( sD /home/psgendb/BIRCHDEV/install/quast-5.2.0/quast_libs/busco/busco.pyt main st 3" %
$
t __main__( t __doc__R> R_ R\ R? R' t loggingt quast_libs.busco.BuscoConfigR R t os.pathR t quast_libs.qutilsR R5 Rn t __name__( ( ( sD /home/psgendb/BIRCHDEV/install/quast-5.2.0/quast_libs/busco/busco.pyt s j V