--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/snpsift_dbnsfp.py	Thu Jan 22 08:04:59 2015 -0500
@@ -0,0 +1,85 @@
+"""
+SnpSift dbNSFP datatypes
+"""
+import os,os.path,re,sys,gzip,logging
+import traceback
+import galaxy.datatypes.data
+from galaxy.datatypes.data import Text
+from galaxy.datatypes.metadata import MetadataElement
+
+log = logging.getLogger(__name__)
+
+class SnpSiftDbNSFP( Text ):
+    """Class describing a dbNSFP database prepared fpr use by SnpSift dbnsfp """
+    MetadataElement( name='reference_name', default='dbSNFP' , desc='Reference Name', readonly=True, visible=True, set_in_upload=True, no_value='dbSNFP' )
+    MetadataElement( name="bgzip", default=None, desc="dbNSFP bgzip", readonly=True, visible=True, no_value=None )
+    MetadataElement( name="index", default=None, desc="Tabix Index File", readonly=True, visible=True, no_value=None)
+    MetadataElement( name="annotation", default=[], desc="Annotation Names", readonly=True, visible=True, no_value=[] )
+    file_ext = "snpsiftdbnsfp"
+    composite_type = 'auto_primary_file'
+    allow_datatype_change = False
+    """
+    ## The dbNSFP file is a tabular file with 1 header line 
+    ## The first 4 columns are required to be: chrom	pos	ref	alt
+    ## These match columns 1,2,4,5 of the VCF file
+    ## SnpSift requires the file to be block-gzipped and the indexed with samtools tabix
+    ## Example:
+    ## Compress using block-gzip algorithm
+    bgzip dbNSFP2.3.txt
+    ## Create tabix index
+    tabix -s 1 -b 2 -e 2 dbNSFP2.3.txt.gz
+    """
+    def __init__( self, **kwd ):
+        Text.__init__( self, **kwd )
+        self.add_composite_file( '%s.grp', description = 'Group File', substitute_name_with_metadata = 'reference_name', is_binary = False )
+        self.add_composite_file( '%s.ti', description = '', substitute_name_with_metadata = 'reference_name', is_binary = False )
+    def init_meta( self, dataset, copy_from=None ):
+        Text.init_meta( self, dataset, copy_from=copy_from )
+    def generate_primary_file( self, dataset = None ):
+        """ 
+        This is called only at upload to write the html file
+        cannot rename the datasets here - they come with the default unfortunately
+        """
+        regenerate_primary_file( self, dataset)
+    def regenerate_primary_file(self,dataset):
+        """
+        cannot do this until we are setting metadata 
+        """
+        annotations = "dbNSFP Annotations: %s\n" % ','.join(dataset.metadata.annotation)
+        f = open(dataset.file_name,'a')
+        if dataset.metadata.bgzip:
+            bn = dataset.metadata.bgzip
+            f.write(bn)
+            f.write('\n')
+        f.write(annotations)
+        f.close()
+    def set_meta( self, dataset, overwrite=True, **kwd ):
+        try:
+            efp = dataset.extra_files_path
+            if os.path.exists(efp):
+                flist = os.listdir(efp)
+                for i,fname in enumerate(flist):
+                    if fname.endswith('.gz'):
+                        dataset.metadata.bgzip = fname
+                        try:
+                            fh = gzip.open(os.path.join(efp,fname),'r')
+                            buf = fh.read(5000)
+                            lines = buf.splitlines()
+                            headers = lines[0].split('\t')
+                            dataset.metadata.annotation = headers[4:]
+                        except Exception,e:        
+                            log.warn("set_meta fname: %s  %s" % (fname,str(e)))
+                            traceback.print_stack(file=sys.stderr)
+                        finally:
+                            fh.close()
+                    if fname.endswith('.tbi'):
+                        dataset.metadata.index = fname
+            self.regenerate_primary_file(dataset)
+        except Exception,e:
+            log.warn("set_meta fname: %s  %s" % (dataset.file_name if dataset and dataset.file_name else 'Unkwown',str(e)))
+            traceback.print_stack(file=sys.stderr)
+
+if __name__ == '__main__':
+    import doctest
+    doctest.testmod(sys.modules[__name__])
+