[MRG] add citation output

CITATION.cff

@@ -0,0 +1,13 @@
+cff-version: 1.0.3
+message: If you use this software, please cite it as below.
+title: sourmash: a library for MinHash sketching of DNA
+version: 2.0.0
+doi: 10.21105/joss.00027
+date-released: 2019-01-10
+authors:
+  - family-names: Brown
+    given-names: C. Titus
+    orcid: https://orcid.org/0000-0001-6001-2677
+  - family-names: Irber Junior
+    given-names: Luiz Carlos
+    orcid: https://orcid.org/0000-0003-4371-9659

doc/index.rst

@@ -33,6 +33,12 @@ We have two tutorials available: one on `genome and metagenome searching <tutori
 To use sourmash, you must be comfortable with the UNIX command line;
 programmers may find the Python library and API useful as well.
 
+If you use sourmash, please cite us!
+
+   Brown and Irber (2016),
+   **sourmash: a library for MinHash sketching of DNA**
+   Journal of Open Source Software, 1(5), 27, `doi:10.21105/joss.00027 <https://joss.theoj.org/papers/3d793c6e7db683bee7c03377a4a7f3c9>`__
+
 sourmash in brief
 -----------------
 

sourmash/__main__.py

@@ -48,6 +48,7 @@
 ** Documentation is available at https://sourmash.readthedocs.io/
 '''
 
+
 def main():
     set_quiet(False)
 
@@ -65,7 +66,7 @@ def main():
                 'sig': sig_main,
                 'signature': sig_main}
     parser = argparse.ArgumentParser(
-        description='work with compressed sequence representations')
+        description='work with compressed biological sequence representations')
     parser.add_argument('command', nargs='?')
     args = parser.parse_args(sys.argv[1:2])
 

sourmash/commands.py

@@ -10,6 +10,7 @@
 import random
 
 import screed
+from .sourmash_args import SourmashArgumentParser
 from . import DEFAULT_SEED, MinHash, load_sbt_index, create_sbt_index
 from . import signature as sig
 from . import sourmash_args
@@ -23,11 +24,9 @@
 WATERMARK_SIZE = 10000
 
 
-
-
 def info(args):
     "Report sourmash version + version of installed dependencies."
-    parser = argparse.ArgumentParser()
+    parser = SourmashArgumentParser(no_citation=True)
     parser.add_argument('-v', '--verbose', action='store_true',
                         help='report versions of khmer and screed')
     args = parser.parse_args(args)
@@ -60,7 +59,7 @@ def compute(args):
             => creates one output file file.sig, with all sequences from
                file1.fa and file2.fa combined into one signature.
     """
-    parser = argparse.ArgumentParser()
+    parser = SourmashArgumentParser()
     parser.add_argument('filenames', nargs='+',
                         help='file(s) of sequences')
 
@@ -356,7 +355,7 @@ def compare(args):
     "Compare multiple signature files and create a distance matrix."
     import numpy
 
-    parser = argparse.ArgumentParser()
+    parser = SourmashArgumentParser()
     parser.add_argument('signatures', nargs='+', help='list of signatures')
     parser.add_argument('-o', '--output')
     parser.add_argument('--ignore-abundance', action='store_true',
@@ -499,7 +498,7 @@ def plot(args):
     from . import fig as sourmash_fig
 
     # set up cmd line arguments
-    parser = argparse.ArgumentParser()
+    parser = SourmashArgumentParser()
     parser.add_argument('distances', help="output from 'sourmash compare'")
     parser.add_argument('--pdf', action='store_true',
                         help='output PDF, not PNG.')
@@ -601,7 +600,7 @@ def plot(args):
 
 def import_csv(args):
     "Import a CSV file full of signatures/hashes."
-    p = argparse.ArgumentParser()
+    p = SourmashArgumentParser()
     p.add_argument('mash_csvfile')
     p.add_argument('-o', '--output', type=argparse.FileType('wt'),
                    default=sys.stdout, help='(default: stdout)')
@@ -635,7 +634,7 @@ def import_csv(args):
 
 
 def dump(args):
-    parser = argparse.ArgumentParser()
+    parser = SourmashArgumentParser()
     parser.add_argument('filenames', nargs='+')
     parser.add_argument('-k', '--ksize', type=int, default=DEFAULT_LOAD_K, help='k-mer size (default: %(default)i)')
     args = parser.parse_args(args)
@@ -654,7 +653,7 @@ def dump(args):
 
 
 def sbt_combine(args):
-    parser = argparse.ArgumentParser()
+    parser = SourmashArgumentParser()
     parser.add_argument('sbt_name', help='name to save SBT into')
     parser.add_argument('sbts', nargs='+',
                         help='SBTs to combine to a new SBT')
@@ -683,7 +682,7 @@ def index(args):
     """
     Build an Sequence Bloom Tree index of the given signatures.
     """
-    parser = argparse.ArgumentParser()
+    parser = SourmashArgumentParser()
     parser.add_argument('sbt_name', help='name to save SBT into')
     parser.add_argument('signatures', nargs='+',
                         help='signatures to load into SBT')
@@ -791,7 +790,7 @@ def index(args):
 def search(args):
     from .search import search_databases
 
-    parser = argparse.ArgumentParser()
+    parser = SourmashArgumentParser()
     parser.add_argument('query', help='query signature')
     parser.add_argument('databases', help='signatures/SBTs to search',
                         nargs='+')
@@ -898,7 +897,7 @@ def search(args):
 
 
 def categorize(args):
-    parser = argparse.ArgumentParser()
+    parser = SourmashArgumentParser()
     parser.add_argument('sbt_name', help='name of SBT to load')
     parser.add_argument('queries', nargs='+',
                         help='list of signatures to categorize')
@@ -980,7 +979,7 @@ def categorize(args):
 def gather(args):
     from .search import gather_databases, format_bp
 
-    parser = argparse.ArgumentParser()
+    parser = SourmashArgumentParser()
     parser.add_argument('query', help='query signature')
     parser.add_argument('databases', help='signatures/SBTs to search',
                         nargs='+')
@@ -1120,7 +1119,7 @@ def gather(args):
 def multigather(args):
     from .search import gather_databases, format_bp
 
-    parser = argparse.ArgumentParser()
+    parser = SourmashArgumentParser()
     parser.add_argument('--db', nargs='+', action='append')
     parser.add_argument('--query', nargs='+', action='append')
     parser.add_argument('--traverse-directory', action='store_true',
@@ -1279,7 +1278,7 @@ def multigather(args):
 def watch(args):
     "Build a signature from raw FASTA/FASTQ coming in on stdin, search."
 
-    parser = argparse.ArgumentParser()
+    parser = SourmashArgumentParser()
     parser.add_argument('sbt_name', help='name of SBT to search')
     parser.add_argument('inp_file', nargs='?', default='/dev/stdin')
     parser.add_argument('-q', '--quiet', action='store_true',
@@ -1376,7 +1375,7 @@ def do_search():
 def storage(args):
     from .sbt import convert_cmd
 
-    parser = argparse.ArgumentParser()
+    parser = SourmashArgumentParser()
     parser.add_argument('-q', '--quiet', action='store_true',
                         help='suppress non-error output')
 
@@ -1394,7 +1393,7 @@ def storage(args):
 
 
 def migrate(args):
-    parser = argparse.ArgumentParser()
+    parser = SourmashArgumentParser()
     parser.add_argument('sbt_name', help='name to save SBT into')
 
     args = parser.parse_args(args)

sourmash/lca/command_classify.py

@@ -12,6 +12,7 @@
 from ..logging import notify, error, debug, set_quiet
 from . import lca_utils
 from .lca_utils import check_files_exist
+from ..sourmash_args import SourmashArgumentParser
 
 DEFAULT_THRESHOLD=5                  # how many counts of a taxid at min
 
@@ -79,7 +80,7 @@ def classify(args):
     """
     main single-genome classification function.
     """
-    p = argparse.ArgumentParser(prog="sourmash lca classify")
+    p = SourmashArgumentParser(prog="sourmash lca classify")
     p.add_argument('--db', nargs='+', action='append')
     p.add_argument('--query', nargs='+', action='append')
     p.add_argument('--threshold', type=int, default=DEFAULT_THRESHOLD)

sourmash/lca/command_compare_csv.py

@@ -12,10 +12,11 @@
 from . import lca_utils
 from .lca_utils import zip_lineage
 from .command_index import load_taxonomy_assignments
+from ..sourmash_args import SourmashArgumentParser
 
 
 def compare_csv(args):
-    p = argparse.ArgumentParser(prog="sourmash lca compare_csv")
+    p = SourmashArgumentParser(prog="sourmash lca compare_csv")
     p.add_argument('csv1', help='taxonomy spreadsheet output by classify')
     p.add_argument('csv2', help='custom taxonomy spreadsheet')
     p.add_argument('-q', '--quiet', action='store_true',

sourmash/lca/command_gather.py

@@ -15,6 +15,8 @@
 from . import lca_utils
 from .lca_utils import check_files_exist
 from ..search import format_bp
+from ..sourmash_args import SourmashArgumentParser
+
 
 LCAGatherResult = namedtuple('LCAGatherResult',
                              'intersect_bp, f_unique_to_query, f_unique_weighted, average_abund, lineage, f_match, name, n_equal_matches')
@@ -183,7 +185,7 @@ def gather_main(args):
     full lineage information for each known hash, as opposed to storing only
     the least-common-ancestor information for it.
     """
-    p = argparse.ArgumentParser(prog="sourmash lca gather")
+    p = SourmashArgumentParser(prog="sourmash lca gather")
     p.add_argument('query')
     p.add_argument('db', nargs='+')
     p.add_argument('-o', '--output', type=argparse.FileType('wt'),

sourmash/lca/command_index.py

@@ -12,6 +12,7 @@
 from ..logging import notify, error, debug, set_quiet
 from . import lca_utils
 from .lca_utils import LineagePair
+from ..sourmash_args import SourmashArgumentParser
 
 
 def load_taxonomy_assignments(filename, delimiter=',', start_column=2,
@@ -111,7 +112,7 @@ def index(args):
     """
     main function for building an LCA database.
     """
-    p = argparse.ArgumentParser(prog="sourmash lca index")
+    p = SourmashArgumentParser(prog="sourmash lca index")
     p.add_argument('csv', help='taxonomy spreadsheet')
     p.add_argument('lca_db_out', help='name to save database to')
     p.add_argument('signatures', nargs='+',

sourmash/lca/command_rankinfo.py

@@ -9,6 +9,7 @@
 
 from ..logging import error, debug, set_quiet
 from . import lca_utils
+from ..sourmash_args import SourmashArgumentParser
 
 
 def make_lca_counts(dblist):
@@ -48,7 +49,7 @@ def rankinfo_main(args):
     """
     rankinfo!
     """
-    p = argparse.ArgumentParser(prog="sourmash lca rankinfo")
+    p = SourmashArgumentParser(prog="sourmash lca rankinfo")
     p.add_argument('db', nargs='+')
     p.add_argument('--scaled', type=float)
     p.add_argument('-q', '--quiet', action='store_true',

sourmash/lca/command_summarize.py

@@ -12,6 +12,8 @@
 from ..logging import notify, error, print_results, set_quiet, debug
 from . import lca_utils
 from .lca_utils import check_files_exist
+from ..sourmash_args import SourmashArgumentParser
+
 
 DEFAULT_THRESHOLD=5
 
@@ -58,7 +60,7 @@ def summarize_main(args):
     """
     main summarization function.
     """
-    p = argparse.ArgumentParser(prog="sourmash lca summarize")
+    p = SourmashArgumentParser(prog="sourmash lca summarize")
     p.add_argument('--db', nargs='+', action='append')
     p.add_argument('--query', nargs='+', action='append')
     p.add_argument('--threshold', type=int, default=DEFAULT_THRESHOLD)

sourmash/sig/__main__.py

@@ -12,7 +12,7 @@
 
 from ..logging import set_quiet, error, notify, set_quiet, print_results
 from .. import sourmash_args
-from ..sourmash_args import DEFAULT_LOAD_K
+from ..sourmash_args import DEFAULT_LOAD_K, SourmashArgumentParser
 from .._minhash import get_max_hash_for_scaled
 
 usage='''
@@ -68,7 +68,7 @@ def describe(args):
     """
     provide basic info on signatures
     """
-    p = argparse.ArgumentParser(prog='sourmash signature describe')
+    p = SourmashArgumentParser(prog='sourmash signature describe')
     p.add_argument('signatures', nargs='+')
     p.add_argument('-q', '--quiet', action='store_true',
                    help='suppress non-error output')
@@ -143,7 +143,7 @@ def overlap(args):
     """
     provide detailed comparison of two signatures
     """
-    p = argparse.ArgumentParser(prog='sourmash signature overlap')
+    p = SourmashArgumentParser(prog='sourmash signature overlap')
     p.add_argument('signature1')
     p.add_argument('signature2')
     p.add_argument('-q', '--quiet', action='store_true',
@@ -231,7 +231,7 @@ def merge(args):
     """
     merge one or more signatures.
     """
-    p = argparse.ArgumentParser(prog='sourmash signature merge')
+    p = SourmashArgumentParser(prog='sourmash signature merge')
     p.add_argument('signatures', nargs='+')
     p.add_argument('-q', '--quiet', action='store_true',
                    help='suppress non-error output')
@@ -299,7 +299,7 @@ def intersect(args):
 
     This function always removes abundances.
     """
-    p = argparse.ArgumentParser(prog='sourmash signature intersect')
+    p = SourmashArgumentParser(prog='sourmash signature intersect')
     p.add_argument('signatures', nargs='+')
     p.add_argument('-q', '--quiet', action='store_true',
                    help='suppress non-error output')
@@ -347,7 +347,7 @@ def subtract(args):
     """
     subtract one or more signatures from another
     """
-    p = argparse.ArgumentParser(prog='sourmash signature subtract')
+    p = SourmashArgumentParser(prog='sourmash signature subtract')
     p.add_argument('signature_from')
     p.add_argument('subtraction_sigs', nargs='+')
     p.add_argument('-q', '--quiet', action='store_true',
@@ -409,7 +409,7 @@ def rename(args):
     """
     rename a signature.
     """
-    p = argparse.ArgumentParser(prog='sourmash signature rename')
+    p = SourmashArgumentParser(prog='sourmash signature rename')
     p.add_argument('signature')
     p.add_argument('name')
     p.add_argument('-q', '--quiet', action='store_true',
@@ -434,7 +434,7 @@ def extract(args):
     """
     extract signatures.
     """
-    p = argparse.ArgumentParser(prog='sourmash signature extract')
+    p = SourmashArgumentParser(prog='sourmash signature extract')
     p.add_argument('signatures', nargs='+')
     p.add_argument('-q', '--quiet', action='store_true',
                    help='suppress non-error output')
@@ -486,7 +486,7 @@ def flatten(args):
     """
     flatten a signature, removing abundances.
     """
-    p = argparse.ArgumentParser(prog='sourmash signature flatten')
+    p = SourmashArgumentParser(prog='sourmash signature flatten')
     p.add_argument('signatures', nargs='+')
     p.add_argument('-q', '--quiet', action='store_true',
                    help='suppress non-error output')
@@ -541,7 +541,7 @@ def downsample(args):
     """
     downsample a scaled signature.
     """
-    p = argparse.ArgumentParser(prog='sourmash signature downsample')
+    p = SourmashArgumentParser(prog='sourmash signature downsample')
     p.add_argument('signatures', nargs="+")
     p.add_argument('--scaled', type=int, default=0,
                    help='scaled value to downsample to')
@@ -612,7 +612,7 @@ def sig_import(args):
     """
     import a signature into sourmash format.
     """
-    p = argparse.ArgumentParser(prog='sourmash signature import')
+    p = SourmashArgumentParser(prog='sourmash signature import')
     p.add_argument('filenames', nargs='+')
     p.add_argument('-q', '--quiet', action='store_true',
                    help='suppress non-error output')
@@ -650,7 +650,7 @@ def export(args):
     """
     export a signature to mash format
     """
-    p = argparse.ArgumentParser(prog='sourmash signature export')
+    p = SourmashArgumentParser(prog='sourmash signature export')
     p.add_argument('filename')
     p.add_argument('-q', '--quiet', action='store_true',
                    help='suppress non-error output')