def test_single_row(self, rw_data): myDataList = [] curContainer = DataContainer("myblock") aCat = DataCategory("pdbx_seqtool_mapping_ref") aCat.appendAttribute("ordinal") aCat.appendAttribute("entity_id") aCat.appendAttribute("auth_mon_id") aCat.appendAttribute("auth_mon_num") aCat.appendAttribute("pdb_chain_id") aCat.appendAttribute("ref_mon_id") aCat.appendAttribute("ref_mon_num") aCat.appendAttribute("details") aCat.append([1, 2, 3, 4, 5, 6, 7, 'data_my_big_data_file']) aCat.append([1, 2, 3, 4, 5, 6, 7, 'loop_my_big_data_loop']) aCat.append([1, 2, 3, 4, 5, 6, 7, 'save_my_big_data_saveframe']) aCat.append([1, 2, 3, 4, 5, 6, 7, '_category.item']) curContainer.append(aCat) bCat = curContainer.getObj("pdbx_seqtool_mapping_ref") print("----attribute list %r\n" % bCat.getAttributeList()) row = bCat.getRow(0) print("----ROW %r\n" % row) with open(str(rw_data['pathOutputFile2']), "w") as ofh: myDataList.append(curContainer) pdbxW = PdbxWriter(ofh) pdbxW.write(myDataList) assert len(myDataList) == 1
def testSingleRow(self): """Test case - read /write single row and null row in data file """ try: # myDataList = [] # ofh = open(self.__pathOutputFile1, "w") curContainer = DataContainer("myblock") aCat = DataCategory("pdbx_seqtool_mapping_ref") aCat.appendAttribute("ordinal") aCat.appendAttribute("entity_id") aCat.appendAttribute("auth_mon_id") aCat.appendAttribute("auth_mon_num") aCat.appendAttribute("pdb_chain_id") aCat.appendAttribute("ref_mon_id") aCat.appendAttribute("ref_mon_num") aCat.appendAttribute("details") aCat.append([1, 2, 3, 4, 5, 6, 7, "data_my_big_data_file"]) aCat.append([1, 2, 3, 4, 5, 6, 7, "loop_my_big_data_loop"]) aCat.append([1, 2, 3, 4, 5, 6, 7, "save_my_big_data_saveframe"]) aCat.append([1, 2, 3, 4, 5, 6, 7, "_category.item"]) # aCat.dumpIt() curContainer.append(aCat) # bCat = curContainer.getObj("pdbx_seqtool_mapping_ref") logger.debug("----attribute list %r", bCat.getAttributeList()) row = bCat.getRow(0) logger.debug("----ROW %r", row) # with open(self.__pathOutputFile2, "w") as ofh: myDataList.append(curContainer) pdbxW = PdbxWriter(ofh) pdbxW.write(myDataList) self.assertEqual(len(myDataList), 1) except Exception as e: logger.exception("Failing with %s", str(e)) self.fail()
class CIF(object): """ This class uses the mmcif library to create an mmCIF-like object Each object has one container, with a container ID and a list of DataCategory objects """ DUMMY_CONTAINER_ID = "emd_0000" def __init__(self, cif_name_name): self.filename = cif_name_name # self.__dataList needed for PDBxWriter self.__dataList = [] self.__container_id = None self.__container = None self.__dataMap = {} def write(self): """ Given a file name, a pdbx writer is used to write data stored in self.__dataList :return written: a boolean; True when pdf writer is finished """ written = False if self.filename: ofh = open(self.filename, "w") pdbx_writer = PdbxWriter(ofh) pdbx_writer.write(self.__dataList) ofh.close() written = True return written def add_container(self, container_id): """ This method provides the basic functionality to set up a container :param container_id: a string; an mmcif category e.g. 'emd_admin' :return: """ added = False self.__container_id = container_id self.__container = DataContainer(container_id) self.__dataMap[container_id] = len(self.__dataList) self.__dataList.append(self.__container) if self.__container is not None: added = True return added def prepare_container(self, container_id): """ Creates a container is it doesn't exist using either provided value or the dummy value :param container_id: a string; an mmcif category e.g. 'emd_admin' :return: """ if not self.__container: if container_id is None: container_id = self.DUMMY_CONTAINER_ID return self.add_container(container_id) def add_category(self, category_id, items): """ This method creates a data category object, adds all items to it and appends it to the container :param category_id: a string; an mmcif category e.g. 'emd_admin' :param items: a list of strings; each element in the list is an item of mmcif category as defined by category_id :return: a list of strings; each element represents a value for the corresponding element in data_items """ category = DataCategory(category_id) for item in items: category.appendAttribute(item) self.__container.append(category) # # def update_single_row_value(self, category_id, item_name, row, value): # """Update value in single row # """ # catObj = self.__container.getObj(category_id) # if catObj is None: # return # # # catObj.setValue(value, item_name, row) # # def update_multiple_rows_value(self, category_id, item_name, value): # """Update value in multiple rows # """ # cat_obj = self.__container.getObj(category_id) # if cat_obj is None: # return # # # row_no = cat_obj.getRowCount() # for row in range(0, row_no): # cat_obj.setValue(value, item_name, row) def insert_data(self, category_id, data_list): """ This method appends the data in data_list to the container labeled category_id :param category_id: a string; an mmcif category e.g. 'emd_admin' :param data_list: :return: """ cat_obj = self.__container.getObj(category_id) if cat_obj is None: return if any(isinstance(el, list) for el in data_list): # print(data_list) for data_ord in data_list[0]: new_list = [] ord_index = data_list[0].index(data_ord) new_list.append(ord_index) new_list.append(data_list[1][ord_index]) # print(new_list) cat_obj.append(new_list) else: cat_obj.append(data_list) def insert_data_into_category(self, category_id, data_items, data_list): """ Helper method: calls two other methods, one to add a category and its items into a container and another to insert the data for the category items :param category_id: a string; an mmcif category e.g. 'emd_admin' :param data_items: a list of strings; each element in the list is an item of mmcif category as defined by category_id :param data_list: a list of strings; each element represents a value for the corresponding element in data_items :return: """ # print('INSERT DATA INTO CATEGORY:', category_id, data_items, data_list) self.add_category(category_id, data_items) self.insert_data(category_id, data_list)
def __writeModel(self, targetId, targetObj, fitFD, fitXyzMapD, fitAtomUnMappedL, matchD, modelId, modelPath): """Write the chemical component model for the input chemical component Id and associated atom mapping and feature details -- ComponentAtomDetails = namedtuple("ComponentAtomDetails", "index atNo name aType x y z fCharge") AlignAtomMap = namedtuple("AlignAtomMap", "refId refAtIdx refAtNo refAtName fitId fitAtIdx fitAtNo fitAtName") AlignAtomUnMapped = namedtuple("AlignAtomUnMapped", "fitId fitAtIdx fitAtNo fitAtType fitAtName fitAtFormalCharge x y z fitNeighbors") """ try: unMappedTypeD = defaultdict(int) hAtomPrefix = "HEX" variantType = self.__getBuildVariant(targetId) # if not self.__testUnMappedProtonation(fitAtomUnMappedL): logger.info( "Unmapped non-hydrogen atoms target %r model %r unMapped count (%d)", targetId, modelId, len(fitAtomUnMappedL)) return False, variantType # Get atom partners for the unmapped atoms fitAtMapD = {} for refAtName, fAtTup in fitXyzMapD.items(): fitAtMapD[fAtTup.atName] = refAtName if fitAtomUnMappedL: # Check if neighbors are all mapped ok = True for fitUnTup in fitAtomUnMappedL: for nAtName in fitUnTup.fitNeighbors: if nAtName not in fitAtMapD: ok = False logger.info( "Missing mapped neighbor for %r target %r model %r", nAtName, targetId, modelId) break if not ok: return False, variantType else: logger.debug("%s match has unmapped protonation", modelId) variantType = "tautomer_protomer" # # kList = ["xyz", "SMILES", "SMILES_STEREO", "InChI", "InChIKey"] for k in kList: if k not in fitFD: logger.error( "Fit feature dictionary for %s missing key %s", targetId, k) return False, variantType # ------------ dataContainer = DataContainer(modelId) # myContainer = targetObj dbName = myContainer.getName() if dbName.upper() != targetId.upper(): logger.info("mismatch datablock (%r) and targetId (%r)", dbName, targetId) cObj = None if myContainer.exists("chem_comp"): cObj = myContainer.getObj("chem_comp") # # catName = "pdbx_chem_comp_model" if not dataContainer.exists(catName): dataContainer.append( DataCategory(catName, attributeNameList=["id", "comp_id"])) # parentId = targetId.split("|")[0] wObj = dataContainer.getObj(catName) wObj.setValue(modelId, "id", 0) wObj.setValue(parentId, "comp_id", 0) # # -------- --------- catName = "pdbx_chem_comp_model_atom" if not dataContainer.exists(catName): dataContainer.append( DataCategory(catName, attributeNameList=[ "model_id", "atom_id", "type_symbol", "charge", "model_Cartn_x", "model_Cartn_y", "model_Cartn_z", "ordinal_id" ])) wObj = dataContainer.getObj(catName) # if myContainer.exists("chem_comp_atom"): cObj = myContainer.getObj("chem_comp_atom") # # Only write the mapped atoms in case we are missing hydrogens in the mapping # jj = 0 for ii in range(cObj.getRowCount()): atName = cObj.getValue("atom_id", ii) atType = cObj.getValue("type_symbol", ii) if atName not in fitXyzMapD: unMappedTypeD[atType] += 1 continue fitXyz = fitXyzMapD[atName] # # fCharge = cObj.getValue("charge", ii) # wObj.setValue(modelId, "model_id", jj) wObj.setValue(atName, "atom_id", jj) wObj.setValue(atType, "type_symbol", jj) # wObj.setValue(fitXyz.atFormalCharge, "charge", jj) wObj.setValue("%.4f" % fitXyz.x, "model_Cartn_x", jj) wObj.setValue("%.4f" % fitXyz.y, "model_Cartn_y", jj) wObj.setValue("%.4f" % fitXyz.z, "model_Cartn_z", jj) wObj.setValue(jj + 1, "ordinal_id", jj) jj += 1 # # Add the unmapped atoms ... # AlignAtomUnMapped = namedtuple("AlignAtomUnMapped", "fitId fitAtIdx fitAtNo fitAtType fitAtName fitNeighbors") ii = wObj.getRowCount() for jj, uTup in enumerate(fitAtomUnMappedL): refAtomName = hAtomPrefix + str(jj) wObj.setValue(modelId, "model_id", ii) wObj.setValue(refAtomName, "atom_id", ii) wObj.setValue(uTup.fitAtType, "type_symbol", ii) wObj.setValue(uTup.fitAtFormalCharge, "charge", ii) wObj.setValue("%.4f" % uTup.x, "model_Cartn_x", ii) wObj.setValue("%.4f" % uTup.y, "model_Cartn_y", ii) wObj.setValue("%.4f" % uTup.z, "model_Cartn_z", ii) wObj.setValue(ii + 1, "ordinal_id", ii) # -------- --------- catName = "pdbx_chem_comp_model_bond" if not dataContainer.exists(catName): dataContainer.append( DataCategory(catName, attributeNameList=[ "model_id", "atom_id_1", "atom_id_2", "value_order", "ordinal_id" ])) wObj = dataContainer.getObj(catName) # if myContainer.exists("chem_comp_bond"): cObj = myContainer.getObj("chem_comp_bond") # jj = 0 for ii in range(cObj.getRowCount()): at1 = cObj.getValue("atom_id_1", ii) if at1 not in fitXyzMapD: continue at2 = cObj.getValue("atom_id_2", ii) if at2 not in fitXyzMapD: continue bType = cObj.getValue("value_order", ii) # wObj.setValue(modelId, "model_id", jj) wObj.setValue(at1, "atom_id_1", jj) wObj.setValue(at2, "atom_id_2", jj) wObj.setValue(bType, "value_order", jj) wObj.setValue(jj + 1, "ordinal_id", jj) jj += 1 # ii = wObj.getRowCount() for jj, uTup in enumerate(fitAtomUnMappedL): at1 = hAtomPrefix + str(jj) for nAt in uTup.fitNeighbors: at2 = fitAtMapD[nAt] wObj.setValue(modelId, "model_id", ii) wObj.setValue(at1, "atom_id_1", ii) wObj.setValue(at2, "atom_id_2", ii) wObj.setValue("SING", "value_order", ii) wObj.setValue(ii + 1, "ordinal_id", ii) # -------- --------- catName = "pdbx_chem_comp_model_descriptor" if not dataContainer.exists(catName): dataContainer.append( DataCategory( catName, attributeNameList=["model_id", "type", "descriptor"])) wObj = dataContainer.getObj(catName) # ii = 0 wObj.setValue(modelId, "model_id", ii) wObj.setValue("SMILES", "type", ii) wObj.setValue(fitFD["SMILES"], "descriptor", ii) ii += 1 wObj.setValue(modelId, "model_id", ii) wObj.setValue("SMILES_CANONICAL", "type", ii) wObj.setValue(fitFD["SMILES_STEREO"], "descriptor", ii) ii += 1 wObj.setValue(modelId, "model_id", ii) wObj.setValue("InChI", "type", ii) wObj.setValue(fitFD["InChI"], "descriptor", ii) ii += 1 wObj.setValue(modelId, "model_id", ii) wObj.setValue("InChIKey", "type", ii) wObj.setValue(fitFD["InChIKey"], "descriptor", ii) # # -------- --------- if matchD["queryId"] is not None: catName = "pdbx_chem_comp_model_reference" if not dataContainer.exists(catName): dataContainer.append( DataCategory(catName, attributeNameList=[ "model_id", "db_name", "db_code" ])) wObj = dataContainer.getObj(catName) ii = 0 wObj.setValue(modelId, "model_id", ii) wObj.setValue("COD", "db_name", ii) wObj.setValue(matchD["queryId"], "db_code", ii) # featureD = {} v = matchD["rValue"] vS = str(v) if v is not None and len(vS) > 0: featureD["r_factor"] = "%.3f" % float(v) # v = matchD["diffrnTemp"] vS = str(v) # remove string artifacts from temperature string ... if v is not None and len(vS) > 0: tV = vS.upper() try: if tV.endswith("DEG.C"): tV = tV.replace("AT", "") tV = tV.replace("DEG.C", "") tV = float(tV.strip()) tV = tV + 273.15 else: tV = tV.replace("AT", "") tV = tV.replace("K", "") tV = float(tV.strip()) featureD["experiment_temperature"] = tV except Exception as e: logger.exception( "Temperature conversion fails for %s (%r) with %s", modelId, vS, tV) # v = matchD["publicationDOI"] vS = str(v) if v is not None and len(vS) > 0: featureD["publication_doi"] = v # v = matchD["version"] vS = str(v) if v is not None and len(vS) > 0: featureD["cod_version"] = v # if matchD["radiationSource"] and "neutron" in matchD[ "radiationSource"]: featureD["neutron_radiation_experiment"] = True if matchD["hasDisorder"] in ["Y"]: featureD["has_disorder"] = True # if len(unMappedTypeD) == 1 and "H" in unMappedTypeD: logger.info("model %r heavy_atoms_only", modelId) featureD["heavy_atoms_only"] = True else: featureD["all_atoms_have_sites"] = True # -------- --------- catName = "pdbx_chem_comp_model_feature" if not dataContainer.exists(catName): dataContainer.append( DataCategory(catName, attributeNameList=[ "model_id", "feature_name", "feature_value" ])) wObj = dataContainer.getObj(catName) # fKeyList = [ "experiment_temperature", "publication_doi", "r_factor", "csd_version" ] ii = 0 for fKey in fKeyList: if fKey in featureD: wObj.setValue(modelId, "model_id", ii) wObj.setValue(fKey, "feature_name", ii) wObj.setValue(str(featureD[fKey]), "feature_value", ii) ii += 1 # boolKeyList = [ "has_disorder", "neutron_radiation_experiment", "heavy_atoms_only", "all_atoms_have_sites" ] for fKey in boolKeyList: if fKey in featureD: if featureD[fKey]: wObj.setValue(modelId, "model_id", ii) wObj.setValue(fKey, "feature_name", ii) wObj.setValue("Y", "feature_value", ii) ii += 1 # if variantType: wObj.setValue(modelId, "model_id", ii) wObj.setValue(variantType + "_match", "feature_name", ii) wObj.setValue("Y", "feature_value", ii) ii += 1 # -------- --------- catName = "pdbx_chem_comp_model_audit" if not dataContainer.exists(catName): dataContainer.append( DataCategory( catName, attributeNameList=["model_id", "action_type", "date"])) wObj = dataContainer.getObj(catName) # ii = 0 wObj.setValue(modelId, "model_id", ii) wObj.setValue("Initial release", "action_type", ii) wObj.setValue(self.__getToday(), "date", ii) # wObj.setValue('RCSB', 'processing_site', ii) # wObj.setValue('JDW', 'annotator', ii) # wObj.setValue('?', 'details', ii) # mU = MarshalUtil(workPath=self.__cachePath) ok = mU.doExport(modelPath, [dataContainer], fmt="mmcif") return ok, variantType except Exception as e: logger.exception("Failing for %r with %s", targetId, str(e)) return False, ""
def __buildCif(self, rD, containerName="vrpt"): """ Construct a mmCIF data category objects for the input extracted data. Args: rD (dict): extracted data organized by category. containerName (str) : data block name Returns: containers (list): data container list """ # curContainer = DataContainer(containerName) for elName in rD: catName = elName if (not rD[elName]) or (not self.__attribD[catName]) or ( catName in ["programs"]): continue hasOrdinal = "ordinal" in self.__attribD[catName] rowList = rD[elName] # Find the unique attribute content across the rowlist and the ordinal value atS = set() for ii, rowD in enumerate(rowList, 1): if hasOrdinal: rowD["ordinal"] = ii if "icode" in rowD: rowD["icode"] = str(rowD["icode"]).strip() if "altcode" in rowD: rowD["altcode"] = str(rowD["altcode"]).strip() atS.update(rowD.keys()) attributeNameList = list(atS) # # Set a reasonable order for these attributes # sD = {ky: self.__atOrdD[ky] for ky in attributeNameList} srtAtL = [ tup[0] for tup in sorted(sD.items(), key=operator.itemgetter(1)) ] logger.debug("Category %s sorted attributes %r", catName, srtAtL) aCat = DataCategory(catName, srtAtL, rowList) curContainer.append(aCat) # # Adjust schema names - # atD = self.__dictionaryMap["attributes"] for catName in curContainer.getObjNameList(): catObj = curContainer.getObj(catName) atNameList = catObj.getAttributeList() mapD = {} mapCatName = self.__dictionaryMap["categories"][ catName] if catName in self.__dictionaryMap[ "categories"] else catName for atName in atNameList: mapD[atName] = atD[(catName, atName)]["at"] if ( catName, atName) in atD else atName catObj.renameAttributes(mapD) catObj.setName(mapCatName) # # Map provenance items from programs.properties - # catObj = curContainer.getObj("program") if catObj and catObj.hasAttribute("properties"): for iRow in range(catObj.getRowCount()): pV = catObj.getValue("properties", iRow) pVL = [v.strip() for v in pV.split(",")] nL = [ self.__atMap[ky] if ky in self.__atMap else ky for ky in pVL ] catObj.setValue(",".join(nL), "properties", iRow) # logger.info("Row %r properties %r" % (iRow, pV)) # return [curContainer]
class mmCIFUtil: """Using pdbx mmCIF utility to parse mmCIF file""" def __init__(self, verbose=False, log=sys.stderr, filePath=None): # pylint: disable=unused-argument # self.__verbose = verbose self.__lfh = log self.__filePath = filePath self.__dataList = [] self.__dataMap = {} self.__container = None self.__blockID = None self.__read() # def __read(self): if not self.__filePath: return # try: ifh = open(self.__filePath, "r") pRd = PdbxReader(ifh) pRd.read(self.__dataList) ifh.close() if self.__dataList: self.__container = self.__dataList[0] self.__blockID = self.__container.getName() idx = 0 for container in self.__dataList: self.__dataMap[container.getName()] = idx idx += 1 # # except Exception as e: self.__lfh.write("Read %s failed %s.\n" % (self.__filePath, str(e))) # def GetBlockID(self): """Return first block ID""" return self.__blockID def GetValueAndItemByBlock(self, blockName, catName): """Get category values and item names""" dList = [] iList = [] if blockName not in self.__dataMap: return dList, iList # catObj = self.__dataList[self.__dataMap[blockName]].getObj(catName) if not catObj: return dList, iList # iList = catObj.getAttributeList() rowList = catObj.getRowList() for row in rowList: tD = {} for idxIt, itName in enumerate(iList): if row[idxIt] != "?" and row[idxIt] != ".": tD[itName] = row[idxIt] # if tD: dList.append(tD) # # return dList, iList def GetValueAndItem(self, catName): dList, iList = self.GetValueAndItemByBlock(self.__blockID, catName) return dList, iList def GetValue(self, catName): """Get category values based on category name 'catName'. The results are stored in a list of dictionaries with item name as key """ dList, _iList = self.GetValueAndItemByBlock(self.__blockID, catName) return dList def GetSingleValue(self, catName, itemName): """Get the first value of item name 'itemName' from 'itemName' item in 'catName' category.""" text = "" dlist = self.GetValue(catName) if dlist: if itemName in dlist[0]: text = dlist[0][itemName] return text # def UpdateSingleRowValue(self, catName, itemName, row, value): """Update value in single row""" catObj = self.__container.getObj(catName) if catObj is None: return # catObj.setValue(value, itemName, row) def UpdateMultipleRowsValue(self, catName, itemName, value): """Update value in multiple rows""" catObj = self.__container.getObj(catName) if catObj is None: return # rowNo = catObj.getRowCount() for row in range(0, rowNo): catObj.setValue(value, itemName, row) # def AddBlock(self, blockID): """Add Data Block""" self.__container = DataContainer(blockID) self.__blockID = blockID self.__dataMap[blockID] = len(self.__dataList) self.__dataList.append(self.__container) def AddCategory(self, categoryID, items): """Add Category""" category = DataCategory(categoryID) for item in items: category.appendAttribute(item) # self.__container.append(category) def RemoveCategory(self, categoryID): return self.__container.remove(categoryID) def InsertData(self, categoryID, dataList): """""" catObj = self.__container.getObj(categoryID) if catObj is None: return # for data in dataList: catObj.append(data) # def WriteCif(self, outputFilePath=None): """Write out cif file""" if not outputFilePath: return # ofh = open(outputFilePath, "w") pdbxW = PdbxWriter(ofh) pdbxW.write(self.__dataList) ofh.close() def GetCategories(self): return self.__container.getObjNameList() def GetAttributes(self, category): return self.__container.getObj(category).getAttributeList() def category_as_dict(self, category, block=None): if block is None: block = self.__blockID values, attributes = self.GetValueAndItemByBlock(block, category) data = [[x[y] if y in x else None for y in attributes] for x in values] return {category: {"Items": attributes, "Values": data}} def block_as_dict(self, block=None): if block is None: block = self.__blockID data = {} for category in self.GetCategories(): data.update(self.category_as_dict(category, block=block)) return data