changes to novel precursor code and others

CHANGES

@@ -2,6 +2,19 @@
 CHANGES
 =======
 
+------
+0.5.0
+------
+
+fixes to work with python 3 properly
+snap now works
+updated to use mirbase release 22
+mpl tk fix
+sphinx docs added
+can run novel predictions in parallel
+some fixes to precursor generation
+removed bedtools dependency
+
 ------
 0.4.0
 ------

doc/source/cli.rst

@@ -34,6 +34,8 @@ The advantage of configuration files is in avoiding long commands that have to b
   species = hsa
   pad5 = 3
   pad3 = 5
+  verbose = 1
+  cpus = 1
 
   [aligner]
   default_params = -v 1 --best
@@ -110,6 +112,12 @@ Settings explained:
 | pad3          | 5                               | 5’ flanking bases |
 |               |                                 | to add            |
 +---------------+---------------------------------+-------------------+
+| verbose       | 1                               | print extra       |
+|               |                                 | information       |
++---------------+---------------------------------+-------------------+
+| cpus          | 1                               | number of threads |
+|               |                                 | to use            |
++---------------+---------------------------------+-------------------+
 | sample_labels | samplefile.txt                  | csv file with     |
 |               |                                 | sample labels     |
 +---------------+---------------------------------+-------------------+

doc/source/description.rst

@@ -62,18 +62,14 @@ Linux
 
 On most linux operating systems installations of Python should include the pip tool.
 If not use your distributions package manager to install pip first. Then the simple
-call below should install all dependencies. However if this fails see the linux section
-below for commands to run for installing any pre-requisites that might not be on your
-system.
-::
-  pip install smallrnaseq
+call below should install all dependencies.
 
-If pip fails you can run the following commands first to fix likely missing packages.
-These are mainly needed for HTSeq to install. Then run pip again.
+You should first run the following commands for installing the pre-requisites that might not
+be on your system already.
 
 Ubuntu::
 
-    sudo apt install python-dev samtools bedtools liblzma-dev libbz2-dev zlib1g-dev liblzo2-dev python-scipy
+    sudo apt install python-dev samtools liblzma-dev libbz2-dev zlib1g-dev liblzo2-dev python-scipy
     sudo pip install smallrnaseq
     sudo apt install bowtie
 
@@ -84,6 +80,10 @@ Fedora::
     sudo pip install smallrnaseq
     sudo dnf install bowtie
 
+Then finally run::
+
+  pip install smallrnaseq
+
 Snap package
 ++++++++++++
 
@@ -139,10 +139,10 @@ will install the older version which should work.
 Windows
 -------
 
-In theory this package will work on Windows but has not been tested. If you are a windows user there are several options available:
+If you are a Windows user there are several options available:
 
- 1. simply use linux running inside a virtualbox instance. see http://www.makeuseof.com/tag/how-to-use-virtualbox/
- 2. install windows subsystem for linux (WSL_) and use the linux instructions above (not tested).
+ 1. Simply use linux running inside a virtualbox instance. see http://www.makeuseof.com/tag/how-to-use-virtualbox/
+ 2. Install Windows Subsystem for Linux (WSL_) and use the linux instructions above. This is confirmed to work on Windows 10.
  3. Use conda and bioconda (see the OSX instructions below).
 
 .. _WSL: https://docs.microsoft.com/en-gb/windows/wsl/install-win10
@@ -162,6 +162,8 @@ Required dependencies
 * matplotlib
 * seaborn (requires scipy)
 * HTSeq
+* bx-python
+* pyfaidx
 * scikit-learn
 
 Installing R for differential expression

smallrnaseq/aligners.py

@@ -84,7 +84,7 @@ def build_subread_index(fastafile, path):
     utils.move_files(files, path)
     return
 
-def bowtie_align(infile, ref, outfile=None, remaining=None, verbose=True):
+def bowtie_align(infile, ref, outfile=None, remaining=None, cpus=2, verbose=True):
     """Map reads using bowtie"""
 
     label = os.path.splitext(os.path.basename(infile))[0]
@@ -100,7 +100,7 @@ def bowtie_align(infile, ref, outfile=None, remaining=None, verbose=True):
     params = BOWTIE_PARAMS
     if remaining == None:
         remaining = os.path.join(outpath, label+'_r.fa')
-    cmd = 'bowtie -f -p 2 -S %s --un %s %s %s > %s' %(params,remaining,ref,infile,outfile)
+    cmd = 'bowtie -f -p %s -S %s --un %s %s %s > %s' %(cpus,params,remaining,ref,infile,outfile)
     if verbose == True:
         print (cmd)
     try:

smallrnaseq/app.py

@@ -166,7 +166,8 @@ def map_mirnas(self):
         ref_name = self.ref_name
         mat_name = 'mirbase-%s' %self.species
         self.aligner_params[mat_name] = self.mirna_params
-
+        novel.VERBOSE = self.verbose
+
         if self.check_index(ref_name) == False:
             print ('no index for reference genome')
             ref_name = ''
@@ -213,7 +214,7 @@ def map_mirnas(self):
             new,cl = novel.find_mirnas(allreads, self.ref_fasta, species=self.species,
                                        score_cutoff=float(self.score_cutoff),
                                        read_cutoff=int(self.read_cutoff),
-                                       cpus=self.cpus, verbose=self.verbose)
+                                       cpus=self.cpus)
             if self.strict == True:
                 new = new[new.mature_check=='ok']
                 print ('filtered %s' %len(new))

smallrnaseq/base.py

@@ -294,7 +294,7 @@ def deseq_normalize(df):
 
 def map_rnas(files, indexes, outpath, collapse=True, adapters=None, aligner='bowtie',
              norm_method='library', use_remaining=True, overwrite=False,
-             samplelabels=None, params={}, count_method='split', verbose=True):
+             samplelabels=None, params={}, count_method='split', cpus=2, verbose=True):
     """Map reads to one or more gene annotations, assumes adapters are removed.
 
     Args:
@@ -351,7 +351,7 @@ def map_rnas(files, indexes, outpath, collapse=True, adapters=None, aligner='bow
             rem = os.path.join(outpath, filename+'_r.fa')
 
             if aligner == 'bowtie':
-                aligners.bowtie_align(query, idx, outfile=samfile,
+                aligners.bowtie_align(query, idx, outfile=samfile, cpus=cpus,
                                       remaining=rem, verbose=verbose)
             elif aligner == 'subread':
                 aligners.subread_align(query, idx, samfile)

smallrnaseq/data/styles.css

@@ -129,7 +129,7 @@ a:active { color: red; }
 	margin-right: 5px;
 	float: right;
 	height: 650px;
-	width: 20%;
+	width: 25%;
 	position: fixed;
 	overflow:hidden;
 	top: 30;