package for Jbrowse Backend added

Author: Seren

backend/JBrowseDataSource.py

@@ -0,0 +1,130 @@
+'''
+Created on Feb 8, 2011
+
+@author: uemit.seren
+'''
+
+import os
+import simplejson
+import re
+
+class ChromosomeData(object):
+    
+    
+    def __init__(self,data_folder,trackData):
+        self.__lazyArrayChunks = {}
+        self.__data_folder = data_folder
+        self.__featureNCList = trackData['featureNCList']
+        self.__sublistIndex = trackData['sublistIndex']
+        self.__lazyIndex = trackData['lazyIndex']
+        self.__subFeatureArray = trackData['subfeatureArray']
+        self.__lazyFeatureFile = self.__data_folder + '/' + re.sub('{chunk}','%s',trackData['lazyfeatureUrlTemplate'])
+        self.__subFeatureFile = self.__data_folder+'/'+self.__subFeatureArray['urlTemplate'].replace('{chunk}','%s')
+
+    def getGenes(self,start,end,getFeatures=True):
+        genes = []
+        genes = self._getGenesFromNCList(self.__featureNCList,start, end,getFeatures,genes)
+        return genes
+    
+    def _getGenesFromNCList(self,nclist,start,end,getFeatures = True,genes = []):
+        length = len(nclist)
+        i = self._binary_search(nclist,start)
+        while ((i < length) and (i >= 0) and (nclist[i][0] < end)):
+            if (isinstance(nclist[i][self.__lazyIndex],dict)):
+                fp = open(self.__lazyFeatureFile % nclist[i][self.__lazyIndex]['chunk'])
+                lazyFeatures = simplejson.load(fp)
+                fp.close()
+                genes = self._getGenesFromNCList(lazyFeatures,start,end,getFeatures,genes)
+            else:
+                gene = self._getGeneFeaturesForGene(nclist[i],getFeatures)
+                genes.append(gene)
+            if len(nclist[i]) >= self.__sublistIndex +1 and nclist[i][self.__sublistIndex] != None:
+                genes = self._getGenesFromNCList(nclist[i][self.__sublistIndex],start,end,getFeatures,genes)
+            i = i+1
+        return genes
+    
+    def _getGeneFeaturesForGene(self,gene,getFeatures=True):
+        stripped_gene = gene[0:4]
+        stripped_gene.append([])
+        if gene[4] != None and getFeatures:
+            for subFeature in gene[4]:
+                stripped_gene[4].append(self._getGeneFeaturesFromPos(subFeature,subFeature))
+        return stripped_gene
+    
+    def _getGeneFeaturesFromPos(self,start,end):
+        import math
+        features = []
+        start = max([0,start])
+        end = min([end,self.__subFeatureArray['length']])
+        firstChunk = int(math.floor(start / self.__subFeatureArray['chunkSize']))
+        lastChunk = int(math.floor(end / self.__subFeatureArray['chunkSize']))
+
+        for chunk in range(firstChunk,lastChunk+1):
+            if not chunk in self.__lazyArrayChunks:
+                fp = open(self.__subFeatureFile % chunk)
+                lazyFeatures = simplejson.load(fp)
+                fp.close()
+                self.__lazyArrayChunks[chunk] = lazyFeatures
+            features+=self._getGeneFeaturesFromChunk(chunk,start,end)
+        return features
+    
+    def _getGeneFeaturesFromChunk(self,chunk,start,end):
+        features = []
+        chunkSize = self.__subFeatureArray['chunkSize']
+        firstIndex = chunk*chunkSize
+        chunkStart = max([start - firstIndex,0])
+        chunkEnd = min([end - firstIndex,chunkSize-1])
+        for i in range(chunkStart,chunkEnd+1):
+            features+=self.__lazyArrayChunks[chunk][i]
+        return features
+    
+    @classmethod
+    def _binary_search(cls,arr, item, low=-1, high=None,index =1):
+        if high is None:
+            high = len(arr)
+        while (high - low > 1):
+            mid = (low + high) >> 1 
+            midval = arr[mid][index] 
+            if  midval > item:
+                high = mid
+            elif midval < item : 
+                low = mid
+           
+        return high
+        
+class DataSource(object):
+    '''
+    classdocs
+    '''
+    
+
+    def __init__(self,jbrowse_tracks_folder,track_key):
+        '''
+        Constructor
+        '''
+        self.__jbrowse_tracks_folder = jbrowse_tracks_folder
+        self.__track_key = track_key
+        self.__chromosomeSources = {}
+        dirList=os.listdir(self.__jbrowse_tracks_folder+"/data/tracks/")
+        for fname in dirList:
+            self.__initChromocomeSources(fname)
+    
+    def getGenes(self,chromosome,start,end,getFeatures=True):
+        if chromosome not in self.__chromosomeSources:
+            raise Exception('Chromosome Data-Source %s not found' % chromosome)
+        genes = self.__chromosomeSources[chromosome].getGenes(start,end,getFeatures)
+        return genes
+    
+    
+    def __initChromocomeSources(self,fname):
+        fp = open(self._getChromosomeTrackFolder(fname)+'trackData.json')
+        trackData = simplejson.load(fp)
+        fp.close()
+        self.__chromosomeSources[fname] = ChromosomeData(self.__jbrowse_tracks_folder,trackData)
+    
+    def _getChromosomeTrackFolder(self,chromosome):
+        return self.__jbrowse_tracks_folder + '/data/tracks/%s/%s/' % (chromosome,self.__track_key)
+    
+
+    
+