Merge branch 'develop'

.bumpversion.cfg

@@ -1,5 +1,5 @@
 [bumpversion]
-current_version = 0.3.1
+current_version = 0.3.2
 
 [bumpversion:file:setup.py]
 

CONTRIBUTING.rst

@@ -15,7 +15,7 @@ Types of Contributions
 Report Bugs
 ~~~~~~~~~~~
 
-Report bugs at https://github.com/jrderuiter/imfusion/issues.
+Report bugs at https://github.com/nki-ccb/imfusion/issues.
 
 If you are reporting a bug, please include:
 
@@ -46,7 +46,7 @@ Submit Feedback
 ~~~~~~~~~~~~~~~
 
 The best way to send feedback is to file an issue at
-https://github.com/jrderuiter/imfusion/issues.
+https://github.com/nki-ccb/imfusion/issues.
 
 If you are proposing a feature:
 
@@ -105,5 +105,5 @@ Before you submit a pull request, check that it meets these guidelines:
    your new functionality into a function with a docstring, and add the
    feature to the list in README.rst.
 3. The pull request should work for Python 2.7, 3.4 and 3.5. Check
-   https://travis-ci.org/jrderuiter/imfusion/pull_requests
+   https://travis-ci.org/nki-ccb/imfusion/pull_requests
    and make sure that the tests pass for all supported Python versions.

HISTORY.rst

@@ -2,6 +2,13 @@
 History
 =======
 
+0.3.2 (2017-05-11)
+------------------
+
+* Properly added star-fusion support to star aligner (was previously not
+  fully merged).
+* Changed documentation URLs to new repository.
+
 0.3.1 (2017-05-09)
 ------------------
 

Makefile

@@ -82,7 +82,7 @@ dist: clean ## builds source and wheel package
 	python setup.py sdist bdist_wheel
 
 conda: clean-pyc ## build a conda release
-	conda build --python 3.5 -c bioconda -c r -c jrderuiter conda
+	conda build --python 3.5 -c bioconda conda
 
 conda-docker: clean-pyc
 	docker run -v `pwd`:/imfusion -t -i condaforge/linux-anvil /bin/sh -c 'cd /imfusion && ./scripts/conda_build_docker.sh'

README.rst

@@ -1,9 +1,3 @@
-.. image:: https://img.shields.io/travis/jrderuiter/imfusion/develop.svg
-    :target: https://travis-ci.org/jrderuiter/imfusion
-
-.. image:: https://img.shields.io/coveralls/jrderuiter/imfusion/develop.svg
-    :target: https://coveralls.io/github/jrderuiter/imfusion
-
 IM-Fusion
 =========
 
@@ -42,12 +36,14 @@ Documentation
 =============
 
 IM-Fusion's documentation is available at
-`jrderuiter.github.io/imfusion <http://jrderuiter.github.io/imfusion/>`_.
+`nki-ccb.github.io/imfusion <http://nki-ccb.github.io/imfusion>`_.
 
 References
 ==========
-de Ruiter, JR. *et al.*, 2017. **"Identifying transposon insertions and
-their effects from RNA-sequencing data"** (*Under revision*).
+
+de Ruiter J.R., Kas S.M. *et al.* **"Identifying transposon insertions and their
+effects from RNA-sequencing data"** Nucleic Acids Research 2017, *in press*.
+
 
 License
 =======

conda/meta.yaml

@@ -1,6 +1,6 @@
 package:
   name: imfusion
-  version: 0.3.1
+  version: 0.3.2
 
 build:
   number: 0
@@ -71,7 +71,7 @@ test:
     - featureCounts -v
 
 about:
-  home: https://github.com/jrderuiter/imfusion
+  home: https://github.com/nki-ccb/imfusion
   license: MIT
   summary: "IM-Fusion - Tool for identifying transposon insertions
     and their effects from RNA-sequencing data"

docs/conf.py

@@ -255,8 +255,7 @@
 
 # One entry per manual page. List of tuples
 # (source start file, name, description, authors, manual section).
-man_pages = [(master_doc, 'im-fusion', u'IM-Fusion Documentation', [author],
-              1)]
+man_pages = [(master_doc, 'imfusion', u'IM-Fusion Documentation', [author], 1)]
 
 # If true, show URL addresses after external links.
 #man_show_urls = False

docs/extras.rst

@@ -72,7 +72,7 @@ that this script does expect the required dependencies to be installed
 (FusionFilter, Repeatmasker and Blast).
 
 .. _FusionFilter wiki: https://github.com/FusionFilter/FusionFilter/wiki/Building-a-Custom-FusionFilter-Dataset
-.. _Python script: https://github.com/jrderuiter/imfusion/blob/develop/scripts/starfusion_build_reference.py
+.. _Python script: https://github.com/nki-ccb/imfusion/blob/develop/scripts/starfusion_build_reference.py
 
 Identifying fusions
 ~~~~~~~~~~~~~~~~~~~

docs/getting_started.rst

@@ -1,7 +1,6 @@
 Getting started
 ===============
 
-
 Overview
 --------
 
@@ -39,6 +38,10 @@ supporting code is provided for interactive analyses (such as plotting
 differential expression) and manually running certain steps of the pipeline.
 For more details, see :doc:`api`.
 
+For the STAR aligner, we also support identifying endogenous gene fusions
+using STAR-Fusion. See :doc:`extras` for more details about this type of
+analysis.
+
 Required files
 --------------
 
@@ -65,6 +68,8 @@ be either ‘SD’ or ‘SA’ for splice-donor or splice-acceptor sites respect
 The field may also be left empty for other types of features, however these
 features will not be used by IM-Fusion.
 
+For generating the exon-level expression counts, IM-Fusion needs a
+flattened exon representation of the reference gene features (in GTF format).
 Example reference files
 -----------------------
 

docs/index.rst

@@ -29,9 +29,19 @@ IM-Fusion has the following key features:
   single insertion in a specific sample, or determine the general
   effect of insertions on a given gene within the tumor cohort.
 
+Additionally, by integrating STAR-Fusion into its STAR-based insertion
+identification pipeline, IM-Fusion enables simultaneous identification of
+transposon insertions and endogenous gene fusions. As shown in our manuscript,
+this approach can identify endogenous gene-fusions driving tumorigenesis, which
+are impossible to identify using targeted DNA-based transposon-sequencing
+approaches.
+
+.. _STAR-Fusion: https://github.com/STAR-Fusion/STAR-Fusion/wiki
+
 For more details on the approach and a comparison with existing DNA-sequencing
 approaches, please see our paper **"Identifying transposon insertions and
-their effects from RNA-sequencing data"** (*Currently under revision*).
+their effects from RNA-sequencing data"** (Nucleic Acids Research 2017,
+*in press*).
 
 .. toctree::
    :maxdepth: 2
@@ -41,6 +51,7 @@ their effects from RNA-sequencing data"** (*Currently under revision*).
    installation
    getting_started
    usage
+   extras
    api
    extras
    contributing

docs/usage.rst

@@ -14,7 +14,7 @@ format) into a new Fasta file and builds the indices needed for alignment.
 Separate sub-commands are provided for each supported aligner (currently STAR
 and Tophat-Fusion).
 
-The basic command for building a (STAR-based) reference is as follows:
+The basic command for building a STAR-based reference is as follows:
 
 .. code:: bash
 
@@ -90,7 +90,7 @@ reference, whilst the ``--output_dir`` argument specifies where the
 sample output should be written.
 
 An optional ``--assemble`` argument indicates whether IM-Fusion should perform
-a reference-guided transcript assembly. If given, IM-Fusion runs Stringtie
+a reference-guided transcript assembly. If given, IM-Fusion runs StringTie
 after the RNA-seq alignment to detect novel gene transcripts based on the
 RNA-seq alignment. The results of this assembly are subsequently used in the
 insertion detection step to annotate insertions that involve novel transcripts.
@@ -105,10 +105,11 @@ The command for using Tophat-Fusion is nearly identical:
         --output_dir output/sample_s1 \
         --tophat_threads 4
 
-However, both aligners do have some aligner-specific arguments concerning the
-alignment. See the help of the respective sub-commands for more details. For
-STAR, special attention should be paid to memory usage, as STAR requires
-approximately 30GB of memory (per process) for loading the reference genome.
+However, as was the case when building the reference genomes, both aligners do
+have some aligner-specific arguments concerning the alignment. See the help of
+the respective sub-commands for more details. Again, for STAR special
+attention should be paid to memory usage, as STAR requires approximately
+30GB of memory for loading the reference genome.
 
 Quantifying expression (per sample)
 -----------------------------------

scripts/starfusion_build_reference.py

@@ -21,28 +21,41 @@ def main():
     tmp_dir.mkdir(parents=True, exist_ok=False)
 
     # Filter the patch chromosomes from the gtf, as these are
-    # likely not present in the fasta.
-    logging.info('- Generating cDNA sequences')
+    # likely not present in the fasta. Note we also filter rows without
+    # a transcript_id, as these seem to be problematic for STAR-Fusion.
+    logging.info('- Filtering GTF file')
 
     gtf_path = tmp_dir / 'ref.gtf'
+    tmp_gtf_path = gtf_path.with_suffix('.gtf.tmp')
+
+    with tmp_gtf_path.open('wb') as file_:
+        check_call(
+            ['grep', '-v', r'^\(MG\|JH\|GL\)', str(args.gtf)], stdout=file_)
+
     with gtf_path.open('wb') as file_:
-        check_call(['grep', '-v', r'^\(MG\|JH\|GL\)', str(args.gtf)],
-                   stdout=file_)
+        check_call(['grep', 'transcript_id', str(tmp_gtf_path)], stdout=file_)
+
+    tmp_gtf_path.unlink()
 
     # Create cDNA_seqs file.
+    logging.info('- Generating cDNA sequences')
+
     cdna_path = tmp_dir / 'cDNA_seqs.fa'
     with cdna_path.open('wb') as file_:
         script_path = args.ff_path / 'util' / 'gtf_file_to_cDNA_seqs.pl'
-        check_call(['perl', str(script_path), str(gtf_path), str(args.fasta)],
-                   stdout=file_)
+        check_call(
+            ['perl', str(script_path), str(gtf_path), str(args.fasta)],
+            stdout=file_)
 
     # Build masked cDNA_seqs file using RepeatMasker.
     # Note: requires library to be installed from http://www.girinst.org.
     logging.info('- Masking repeats')
 
     masked_path = cdna_path.with_suffix('.fa.masked')
-    check_call([str(args.rm_path / 'RepeatMasker'), '-pa', str(args.threads), '-s',
-                '-species', 'mouse', '-xsmall', str(cdna_path)])
+    check_call([
+        str(args.rm_path / 'RepeatMasker'), '-pa', str(args.threads), '-s',
+        '-species', 'mouse', '-xsmall', str(cdna_path)
+    ])
 
     # Create blastpairs.
     logging.info('- Creating blast pairs')
@@ -51,37 +64,35 @@ def main():
 
     pair_path = tmp_dir / 'blast_pairs.outfmt6'
     with pair_path.open('wb') as file_:
-        check_call(['blastn',
-                    '-query', str(cdna_path),
-                    '-db', str(masked_path),
-                    '-max_target_seqs', '10000',
-                    '-outfmt', '6',
-                    '-evalue', '1e-3',
-                    '-lcase_masking',
-                    '-num_threads', str(args.threads),
-                    '-word_size', '11'],
-                   stdout=file_)
+        check_call(
+            [
+                'blastn', '-query', str(cdna_path), '-db', str(masked_path),
+                '-max_target_seqs', '10000', '-outfmt', '6', '-evalue', '1e-3',
+                '-lcase_masking', '-num_threads', str(args.threads),
+                '-word_size', '11'
+            ],
+            stdout=file_)
 
     pair_gz_path = pair_path.with_suffix('.gene_syms.outfmt6.gz')
     with gzip.open(str(pair_gz_path), 'wb') as file_:
         script_path = (args.ff_path / 'util' /
                        'blast_outfmt6_replace_trans_id_w_gene_symbol.pl')
-        check_call(['perl', str(script_path), str(cdna_path), str(pair_path)],
-                   stdout=file_)
+        check_call(
+            ['perl', str(script_path), str(cdna_path), str(pair_path)],
+            stdout=file_)
 
     # Prepare library.
     logging.info('- Preparing library')
-    script_path = args.ff_path / 'util' / 'prep_genome_lib.pl'
-    check_call(['perl', str(script_path),
-                '--genome_fa', str(args.fasta),
-                '--gtf', str(gtf_path),
-                '--blast_pairs', str(pair_gz_path),
-                '--cdna_fa', str(cdna_path),
-                '--CPU', str(args.threads),
-                '--max_readlength', str(args.read_length),
-                '--output_dir', str(args.output_dir)])
-
-    # shutil.rmtree(str(args.tmp_dir))
+    script_path = args.ff_path / 'prep_genome_lib.pl'
+    check_call([
+        'perl', str(script_path), '--genome_fa', str(args.fasta), '--gtf',
+        str(gtf_path), '--blast_pairs', str(pair_gz_path), '--cdna_fa',
+        str(cdna_path), '--CPU', str(args.threads), '--max_readlength',
+        str(args.read_length), '--output_dir', str(args.output_dir)
+    ])
+
+    shutil.rmtree(str(args.tmp_dir))
+
 
 def parse_args():
     """Parses command line arguments."""
@@ -91,7 +102,7 @@ def parse_args():
     parser.add_argument('--fasta', required=True, type=Path)
     parser.add_argument('--gtf', required=True, type=Path)
     parser.add_argument('--output_dir', required=True, type=Path)
-    
+
     parser.add_argument('--ff_path', required=False, default='', type=Path)
     parser.add_argument('--rm_path', required=False, default='', type=Path)
 

setup.py

@@ -27,12 +27,12 @@
 
 setuptools.setup(
     name='imfusion',
-    version='0.3.1',
+    version='0.3.2',
     description=('Tool for identifying transposon insertions in '
                  'Insertional Mutagenesis screens from gene-transposon '
                  'fusions using single- and paired-end RNA-sequencing data.'),
     long_description=README + '\n\n' + HISTORY,
-    url='https://github.com/jrderuiter/im-fusion',
+    url='https://github.com/nki-ccb/imfusion',
     author='Julian de Ruiter',
     author_email='[email protected]',
     license='MIT license',

src/imfusion/__init__.py

@@ -2,4 +2,4 @@
 
 __author__ = 'Julian de Ruiter'
 __email__ = '[email protected]'
-__version__ = '0.3.1'
+__version__ = '0.3.2'

src/imfusion/build/indexers/base.py

@@ -249,7 +249,10 @@ def configure_args(cls, parser):
             help='Path to the transposon features (tsv).')
 
         base_group.add_argument(
-            '--output_dir', type=pathlib.Path, required=True)
+            '--output_dir',
+            type=pathlib.Path,
+            required=True,
+            help='Path to write the built reference.')
 
         # Optional blacklist arguments.
         blacklist_group = parser.add_argument_group('Blacklist arguments')

src/imfusion/external/__init__.py

@@ -0,0 +1 @@
+# -*- coding: utf-8 -*-

src/imfusion/external/star_fusion.py

@@ -0,0 +1,24 @@
+"""Module containing functions for calling star-fusion."""
+
+from .util import run_command
+
+
+def star_fusion(junction_path, reference_path, output_dir=None, log_path=None):
+    """Identifies endogenous fusions from an existing STAR alignment.
+
+    Parameters
+    ----------
+    reference_path : pathlib.Path
+        Path to the reference genome.
+    out_base_path : pathlib.Path
+        Base output path for the built index.
+    log_path : pathlib.Path
+        Where to write the log output.
+
+    """
+
+    args = [
+        'STAR-Fusion', '--genome_lib_dir', str(reference_path), '-J',
+        str(junction_path), '--output_dir', str(output_dir)
+    ]
+    run_command(args=args, log_path=log_path)

src/imfusion/insertions/aligners/base.py

@@ -98,7 +98,7 @@ def configure_args(cls, parser):
             type=pathlib.Path,
             required=True,
             help='Path to the index of the augmented reference '
-            'generated by im-fusion build.')
+            'generated by imfusion-build.')
 
         base_group.add_argument(
             '--output_dir',