Mercurial > repos > mvdbeek > mismatch_frequencies

changeset 24:6590be3f8e3f draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
planemo upload for repository https://bitbucket.org/drosofff/gedtools/
author mvdbeek
date Sun, 24 May 2015 11:36:31 -0400
parents ca7b7890ed20
children 1088aadcb5f9
files mismatch_frequencies.py mismatch_frequencies.xml test-data/3mismatches_ago2ip_ovary.bam test-data/3mismatches_ago2ip_s2.bam test-data/mismatch.pdf test-data/mismatch.tab tool_dependencies.xml
diffstat 7 files changed, 204 insertions(+), 68 deletions(-) [+]
line wrap: on
line diff
--- a/mismatch_frequencies.py	Sun May 24 17:33:33 2015 +0200
+++ b/mismatch_frequencies.py	Sun May 24 11:36:31 2015 -0400
@@ -1,16 +1,27 @@
 import pysam, re, string
 import matplotlib.pyplot as plt
 import pandas as pd
+import json
 from collections import defaultdict
 from collections import OrderedDict
 import argparse
+import itertools
 
 class MismatchFrequencies:
     '''Iterate over a SAM/BAM alignment file, collecting reads with mismatches. One
     class instance per alignment file. The result_dict attribute will contain a
     nested dictionary with name, readlength and mismatch count.'''
-    def __init__(self, result_dict={}, alignment_file=None, name="name", minimal_readlength=21, maximal_readlength=21,
-                 number_of_allowed_mismatches=1, ignore_5p_nucleotides=0, ignore_3p_nucleotides=0):
+    def __init__(self, result_dict={}, alignment_file=None, name="name", minimal_readlength=21, 
+                 maximal_readlength=21,
+                 number_of_allowed_mismatches=1, 
+                 ignore_5p_nucleotides=0, 
+                 ignore_3p_nucleotides=0,
+                 possible_mismatches = [
+                        'AC', 'AG', 'AT',
+                        'CA', 'CG', 'CT',
+                        'GA', 'GC', 'GT',
+                        'TA', 'TC', 'TG'
+                ]):
     
         self.result_dict = result_dict
         self.name = name
@@ -19,27 +30,43 @@
         self.number_of_allowed_mismatches = number_of_allowed_mismatches
         self.ignore_5p_nucleotides = ignore_5p_nucleotides
         self.ignore_3p_nucleotides = ignore_3p_nucleotides
+        self.possible_mismatches = possible_mismatches
         
         if alignment_file:
             self.pysam_alignment = pysam.Samfile(alignment_file)
-            result_dict[name]=self.get_mismatches(self.pysam_alignment, minimal_readlength, maximal_readlength)
+            self.references = self.pysam_alignment.references #names of fasta reference sequences
+            result_dict[name]=self.get_mismatches(
+                self.pysam_alignment, 
+                minimal_readlength, 
+                maximal_readlength,
+                possible_mismatches
+            )
     
-    def get_mismatches(self, pysam_alignment, minimal_readlength, maximal_readlength):
+    def get_mismatches(self, pysam_alignment, minimal_readlength, 
+                       maximal_readlength, possible_mismatches):
         mismatch_dict = defaultdict(int)
-        len_dict={}
-        for i in range(minimal_readlength, maximal_readlength+1):
-            len_dict[i]=mismatch_dict.copy()
+        rec_dd = lambda: defaultdict(rec_dd)
+        len_dict = rec_dd()
         for alignedread in pysam_alignment:
             if self.read_is_valid(alignedread, minimal_readlength, maximal_readlength):
-                len_dict[int(alignedread.rlen)]['total_mapped'] += 1
-                MD=alignedread.opt('MD')
+                chromosome = pysam_alignment.getrname(alignedread.rname)
+                try:
+                    len_dict[int(alignedread.rlen)][chromosome]['total valid reads'] += 1
+                except TypeError:
+                    len_dict[int(alignedread.rlen)][chromosome]['total valid reads'] = 1
+                MD = alignedread.opt('MD')
                 if self.read_has_mismatch(alignedread, self.number_of_allowed_mismatches):
                     (ref_base, mismatch_base)=self.read_to_reference_mismatch(MD, alignedread.seq, alignedread.is_reverse)
                     if ref_base == None:
                             continue
                     else:
                         for i, base in enumerate(ref_base):
-                            len_dict[int(alignedread.rlen)][ref_base[i]+' to '+mismatch_base[i]] += 1
+                            if not ref_base[i]+mismatch_base[i] in possible_mismatches:
+                                continue
+                            try:
+                                len_dict[int(alignedread.rlen)][chromosome][ref_base[i]+mismatch_base[i]] += 1
+                            except TypeError:
+                                len_dict[int(alignedread.rlen)][chromosome][ref_base[i]+mismatch_base[i]] = 1
         return len_dict
     
     def read_is_valid(self, read, min_readlength, max_readlength):
@@ -133,6 +160,7 @@
         if is_reverse:
             reference_base=reverseComplement(reference_base)
             mismatched_base=reverseComplement(mismatched_base)
+            mismatch_position=len(readseq)-mismatch_position-1
         if mismatched_base=='N':
             return (None, None)
         if self.mismatch_in_allowed_region(readseq, mismatch_position):
@@ -140,7 +168,6 @@
         else:
             return (None, None)
 
-
 def reverseComplement(sequence):
     '''do a reverse complement of DNA base.
     >>> reverseComplement('ATGC')=='GCAT'
@@ -154,39 +181,66 @@
 def barplot(df, library, axes):
     df.plot(kind='bar', ax=axes, subplots=False,\
             stacked=False, legend='test',\
-            title='Mismatches in TE small RNAs from {0}'.format(library))
-  
-def result_dict_to_df(result_dict):
-    mismatches = []
-    libraries = []
-    for mismatch, library in result_dict.iteritems():
-        mismatches.append(mismatch)
-        libraries.append(pd.DataFrame.from_dict(library, orient='index'))
-    df=pd.concat(libraries, keys=mismatches)
-    df.index.names = ['library', 'readsize']
-    return df
-
+            title='Mismatch frequencies for {0}'.format(library))
+    
 def df_to_tab(df, output):
     df.to_csv(output, sep='\t')
 
-def plot_result(result_dict, args):
-    names=args.name
+def reduce_result(df, possible_mismatches):
+    '''takes a pandas dataframe with full mismatch details and
+    summarises the results for plotting.'''
+    alignments = df['Alignment_file'].unique()
+    readlengths = df['Readlength'].unique()
+    combinations = itertools.product(*[alignments, readlengths]) #generate all possible combinations of readlength and alignment files
+    reduced_dict = {}
+    frames = []
+    last_column = 3+len(possible_mismatches)
+    for combination in combinations:
+        library_subset = df[df['Alignment_file'] == combination[0]]
+        library_readlength_subset = library_subset[library_subset['Readlength'] == combination[1]]
+        sum_of_library_and_readlength = library_readlength_subset.iloc[:,3:last_column+1].sum()
+        if not reduced_dict.has_key(combination[0]):
+            reduced_dict[combination[0]] = {}
+        reduced_dict[combination[0]][combination[1]] = sum_of_library_and_readlength.to_dict()
+    return reduced_dict
+
+def plot_result(reduced_dict, args):
+    names=reduced_dict.keys()
     nrows=len(names)/2+1
     fig = plt.figure(figsize=(16,32))
     for i,library in enumerate (names):
         axes=fig.add_subplot(nrows,2,i+1)
-        library_dict=result_dict[library]
-        for length in library_dict.keys():
-            for mismatch in library_dict[length]:
-                if mismatch == 'total_mapped':
-                    continue
-                library_dict[length][mismatch]=library_dict[length][mismatch]/float(library_dict[length]['total_mapped'])*100
-            del library_dict[length]['total_mapped']
+        library_dict=reduced_dict[library]
         df=pd.DataFrame(library_dict)
+        df.drop(['total aligned reads'], inplace=True)
         barplot(df, library, axes),
-        axes.set_ylabel('Percent of mapped reads with mismatches')
+        axes.set_ylabel('Mismatch count / all valid reads * readlength')
     fig.savefig(args.output_pdf, format='pdf')    
 
+def format_result_dict(result_dict, chromosomes, possible_mismatches):
+    '''Turn nested dictionary into preformatted tab seperated lines'''
+    header = "Reference sequence\tAlignment_file\tReadlength\t" + "\t".join(
+        possible_mismatches) + "\ttotal aligned reads"
+    libraries = result_dict.keys()
+    readlengths = result_dict[libraries[0]].keys()
+    result = []
+    for chromosome in chromosomes:
+        for library in libraries:
+            for readlength in readlengths:
+                line = []
+                line.extend([chromosome, library, readlength])
+                try:
+                    line.extend([result_dict[library][readlength][chromosome].get(mismatch, 0) for mismatch in possible_mismatches])
+                    line.extend([result_dict[library][readlength][chromosome].get(u'total valid reads', 0)])
+                except KeyError:
+                    line.extend([0 for mismatch in possible_mismatches])
+                    line.extend([0])
+                result.append(line)
+    df = pd.DataFrame(result, columns=header.split('\t'))
+    last_column=3+len(possible_mismatches)
+    df['mismatches/per aligned nucleotides'] = df.iloc[:,3:last_column].sum(1)/(df.iloc[:,last_column]*df['Readlength'])
+    return df
+  
 def setup_MismatchFrequencies(args):
     resultDict=OrderedDict()
     kw_list=[{'result_dict' : resultDict, 
@@ -196,20 +250,32 @@
              'maximal_readlength' : args.max,
              'number_of_allowed_mismatches' : args.n_mm,
              'ignore_5p_nucleotides' : args.five_p, 
-             'ignore_3p_nucleotides' : args.three_p} 
+             'ignore_3p_nucleotides' : args.three_p,
+             'possible_mismatches' : args.possible_mismatches }
              for alignment_file, name in zip(args.input, args.name)]
     return (kw_list, resultDict)
 
+def nested_dict_to_df(dictionary):
+    dictionary = {(outerKey, innerKey): values for outerKey, innerDict in dictionary.iteritems() for innerKey, values in innerDict.iteritems()}
+    df=pd.DataFrame.from_dict(dictionary).transpose()
+    df.index.names = ['Library', 'Readlength']
+    return df
+
 def run_MismatchFrequencies(args):
     kw_list, resultDict=setup_MismatchFrequencies(args)
-    [MismatchFrequencies(**kw_dict) for kw_dict in kw_list]
-    return resultDict
+    references = [MismatchFrequencies(**kw_dict).references for kw_dict in kw_list]
+    return (resultDict, references[0])
 
 def main():
-    result_dict=run_MismatchFrequencies(args)
-    df=result_dict_to_df(result_dict)
-    plot_result(result_dict, args)
-    df_to_tab(df, args.output_tab)
+    result_dict, references = run_MismatchFrequencies(args)
+    df = format_result_dict(result_dict, references, args.possible_mismatches)
+    reduced_dict = reduce_result(df, args.possible_mismatches)
+    plot_result(reduced_dict, args)
+    reduced_df = nested_dict_to_df(reduced_dict)
+    df_to_tab(reduced_df, args.output_tab)
+    if not args.expanded_output_tab == None:
+        df_to_tab(df, args.expanded_output_tab)
+    return reduced_dict
 
 if __name__ == "__main__":
     
@@ -218,12 +284,17 @@
     parser.add_argument('--name', nargs='*', help='Name for input file to display in output file. Should have same length as the number of inputs')
     parser.add_argument('--output_pdf', help='Output filename for graph')
     parser.add_argument('--output_tab', help='Output filename for table')
+    parser.add_argument('--expanded_output_tab', default=None, help='Output filename for table')
+    parser.add_argument('--possible_mismatches', default=[
+            'AC', 'AG', 'AT','CA', 'CG', 'CT', 'GA', 'GC', 'GT', 'TA', 'TC', 'TG'
+        ], nargs='+', help='specify mismatches that should be counted for the mismatch frequency. The format is Reference base -> observed base, eg AG for A to G mismatches.')
     parser.add_argument('--min', '--minimal_readlength', type=int, help='minimum readlength')
     parser.add_argument('--max', '--maximal_readlength', type=int, help='maximum readlength')
     parser.add_argument('--n_mm', '--number_allowed_mismatches', type=int, default=1, help='discard reads with more than n mismatches')
     parser.add_argument('--five_p', '--ignore_5p_nucleotides', type=int, default=0, help='when calculating nucleotide mismatch frequencies ignore the first N nucleotides of the read')
     parser.add_argument('--three_p', '--ignore_3p_nucleotides', type=int, default=1, help='when calculating nucleotide mismatch frequencies ignore the last N nucleotides of the read')
-    #args = parser.parse_args(['--input', '3mismatches_ago2ip.bam', '2mismatch.bam', '--name', 'Siomi1', 'Siomi2' , '--five_p', '3','--three_p','3','--output_pdf', 'out.pdf', '--output_tab', 'out.tab', '--min', '21', '--max', '21'])
+    #args = parser.parse_args(['--input', '3mismatches_ago2ip_s2.bam', '3mismatches_ago2ip_ovary.bam','--possible_mismatches','AC','AG', 'CG', 'TG', 'CT','--name', 'Siomi1', 'Siomi2' , '--five_p', '3','--three_p','3','--output_pdf', 'out.pdf', '--output_tab', 'out.tab', '--expanded_output_tab', 'expanded.tab', '--min', '20', '--max', '22'])
     args = parser.parse_args()
-    main()
+    reduced_dict = main()
 
+
--- a/mismatch_frequencies.xml	Sun May 24 17:33:33 2015 +0200
+++ b/mismatch_frequencies.xml	Sun May 24 11:36:31 2015 -0400
@@ -1,11 +1,11 @@
-<tool id="mismatch_frequencies" name="Mismatch Frequencies" version="0.0.3" hidden="false" >
-	<description>Analyze mismatch frequencies in BAM/SAM alignments</description>
-        <requirements>
-    	    <requirement type="package" version="0.7.7">pysam</requirement>
-    	    <requirement type="package" version="0.14">pandas</requirement>
-    	    <requirement type="package" version="1.4">matplotlib</requirement>
-        </requirements>
-	<command interpreter="python">mismatch_frequencies.py --input 
+<tool id="mismatch_frequencies" name="Mismatch Frequencies" version="0.0.9" hidden="false" >
+  <description>Analyze mismatch frequencies in BAM/SAM alignments</description>
+  <requirements>
+    <requirement type="package" version="0.7.7">pysam</requirement>
+    <requirement type="package" version="0.14.1">pandas</requirement>
+    <requirement type="package" version="1.2.1">matplotlib</requirement>
+  </requirements>
+  <command interpreter="python">mismatch_frequencies.py --input 
 		#for i in $rep
 			"$i.input_file" 
 		#end for
@@ -17,20 +17,73 @@
                  --n_mm $number_of_mismatches
                  --five_p $five_p
                  --three_p $three_p
-        </command>
-	<inputs>
-            <repeat name="rep" title="alignment files">
-        	<param name="input_file" type="data" format="bam,sam" label="Alignment file" help="The input alignment file(s) for which to analyze the mismatches."/>
-            </repeat>
-          <param name="number_of_mismatches" label="Maximum number of allowed mismatches per read" help="Discard reads with more than the chosen number of mismatches from the frequency calculation" type="integer" value="3"/>
-          <param name="min_length" label="Minumum read length to analyse" type="integer" value="21"/>
-	  <param name="max_length" label="Maximum read length to analyse" type="integer" value="21"/>
-	  <param name="five_p" label="Ignore mismatches in the first N nucleotides of a read" type="integer" value="0"/>
-	  <param name="three_p" label="Ignore mismatches in the last N nucleotides of a read" help="useful to discriminate between tailing events and editing events" type="integer" value="3"/>
-	</inputs>
-        <outputs>
-                <data format="pdf" name="output_pdf" />
-                <data format="tabular" name="output_tab" />
-        </outputs>
+                 --expanded_output_tab $expanded_tab
+                 --possible_mismatches $possible_mismatches
+  </command>
+  <inputs>
+    <repeat name="rep" title="alignment files">
+      <param name="input_file" type="data" format="bam,sam" label="Alignment file" help="The input alignment file(s) for which to analyze the mismatches."/>
+    </repeat>
+    <param name="number_of_mismatches" label="Maximum number of allowed mismatches per read" help="Discard reads with more than the chosen number of mismatches from the frequency calculation" type="integer" value="3"/>
+    <param name="possible_mismatches" label="Specify mismatches that should be counted" help="Ignores mismatches that are not listed" type="text" value="AC AG AT CA CG CT GA GC GT TA TC TG">
+      <validator type="expression" message="Allowed values are AGCTN, seperated by space.">len([False for char in value if not char in " AGCTN"]) == 0</validator>
+    </param>
+    <param name="min_length" label="Minumum read length to analyse" type="integer" value="21"/>
+    <param name="max_length" label="Maximum read length to analyse" type="integer" value="21"/>
+    <param name="five_p" label="Ignore mismatches in the first N nucleotides of a read" type="integer" value="0"/>
+    <param name="three_p" label="Ignore mismatches in the last N nucleotides of a read" help="useful to discriminate between tailing events and editing events" type="integer" value="3"/>
+    <param help="Output expanded tabular format" label="Nucleotide mismatches per reference sequence" name="expanded" type="select">
+        <option select="true" value="false">No</option>
+        <option value="expanded">Yes</option>
+    </param>
+  </inputs>
+  <outputs>
+    <data format="tabular" name="output_tab" />
+    <data format="fasta" name="expanded_tab">
+        <filter> expanded == "expanded"</filter>
+    </data>
+    <data format="pdf" name="output_pdf" />
+  </outputs>
+  <tests>
+    <test>
+      <param name="rep_0|input_file" value="3mismatches_ago2ip_s2.bam" ftype="bam" />
+      <param name="rep_1|input_file" value="3mismatches_ago2ip_ovary.bam" ftype="bam" />
+      <param name="number_of_mismatches" value="1" />
+      <param name="min_length" value="21" />
+      <param name="max_length" value="21" />
+      <param name="three_p" value="0" />
+      <param name="five_p" value="0" />
+      <output name="tabular" file="mismatch.tab" ftype="tabular"/>
+    <!--
+      <output name="pdf" file="mismatch.pdf" ftype="pdf"/>
+    -->
+    </test>
+  </tests>
+  <help>
 
+.. class:: infomark
+
+
+***What it does***
+
+This tool reconstitues for each aligned read of an alignment file in SAM/BAM format whether
+a mismatch is annotated in the MD tag, and if that is the case counts the identity of the 
+mismatch relative to the reference sequence. The output is a PDF document with the calculated
+frequency for each mismatch that occured relative to the total number of valid reads and a table
+with the corresponding values. Read length can be limited to a specific read length, and 5 prime and 
+3 prime-most nucleotides of a read can be ignored.
+
+----
+
+.. class:: warningmark
+
+***Warning***
+
+This tool skips all read that have insertions and has been tested only with bowtie and bowtie2
+generated alignment files.
+
+Written by Marius van den Beek, m.vandenbeek at gmail . com
+  </help>
+  <citations>
+  </citations>
 </tool>
Binary file test-data/3mismatches_ago2ip_ovary.bam has changed
Binary file test-data/3mismatches_ago2ip_s2.bam has changed
Binary file test-data/mismatch.pdf has changed
--- a/test-data/mismatch.tab	Sun May 24 17:33:33 2015 +0200
+++ b/test-data/mismatch.tab	Sun May 24 11:36:31 2015 -0400
@@ -1,3 +1,3 @@
-library	readsize	A to C	A to G	A to T	C to A	C to G	C to T	G to A	G to C	G to T	T to A	T to C	T to G	total_mapped
-3mismatches_ago2ip_s2.bam	21	31	5484	69	25	40	137	156	109	188	51	196	29	43881
-3mismatches_ago2ip_ovary.bam	21	293	879	411	452	231	872	845	191	473	384	818	324	138649
+Library	Readlength	AC	AG	AT	CA	CG	CT	GA	GC	GT	TA	TC	TG	total aligned reads
+3mismatches_ago2ip_ovary.bam	21	380	1214	524	581	278	1127	1032	239	595	483	973	394	138649
+3mismatches_ago2ip_s2.bam	21	48	6503	106	68	46	173	222	144	220	90	232	40	43881
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/tool_dependencies.xml	Sun May 24 11:36:31 2015 -0400
@@ -0,0 +1,12 @@
+<?xml version="1.0"?>
+<tool_dependency>
+    <package name="pysam" version="0.7.7">
+        <repository changeset_revision="ca10c522f37e" name="package_pysam_0_7_7" owner="iuc" toolshed="https://testtoolshed.g2.bx.psu.edu" />
+    </package>
+    <package name="pandas" version="0.14.1">
+        <repository changeset_revision="e27b7cf19fef" name="package_pandas_0_14" owner="iuc" toolshed="https://testtoolshed.g2.bx.psu.edu" />
+    </package>
+    <package name="matplotlib" version="1.2.1">
+        <repository changeset_revision="dc29f1f60887" name="package_matplotlib_1_2" owner="iuc" toolshed="https://testtoolshed.g2.bx.psu.edu" />
+    </package>
+</tool_dependency>