a#nSc@sdZddlmZddlmZddlmZddlmZddlZyddl Z e j j j dZ e dd kree d d krdZ nd e jjjkre jd d nde jjjkre jdd nde jjjkre jdd ne jjjjZWnek rKZdZ nXdZddZdZddlmZyddlmZdZ Wne!k rdZ nXdZ"e"ej#_$ddZ%dZ&e%ej#_'dZ(e)dZ*dde+dZ,dd d!Z-d"d#de+d$Z.dd%lm/Z/e/j/Z0dd&Z1dd'd(Z2ddd)Z3dd*l4m5Z5dd+lm6Z6dd,e+d-Z7dd.lm8Z8d/Z9dd,d0Z:e;d1krddl<Z<e dkr e<j=d2IJqe j j j dZ e dd krTee d d krTe<j=d3IJqddl>Z>e>j?d4e>j@e>jA\ZBZCeBre<jDeBqndS(5sL Importing pandasTools enables several features that allow for using RDKit molecules as columns of a Pandas dataframe. If the dataframe is containing a molecule format in a column (e.g. smiles), like in this example: >>> from rdkit.Chem import PandasTools >>> import pandas as pd >>> import os >>> from rdkit import RDConfig >>> antibiotics = pd.DataFrame(columns=['Name','Smiles']) >>> antibiotics = antibiotics.append({'Smiles':'CC1(C(N2C(S1)C(C2=O)NC(=O)CC3=CC=CC=C3)C(=O)O)C','Name':'Penicilline G'}, ignore_index=True)#Penicilline G >>> antibiotics = antibiotics.append({'Smiles':'CC1(C2CC3C(C(=O)C(=C(C3(C(=O)C2=C(C4=C1C=CC=C4O)O)O)O)C(=O)N)N(C)C)O','Name':'Tetracycline'}, ignore_index=True)#Tetracycline >>> antibiotics = antibiotics.append({'Smiles':'CC1(C(N2C(S1)C(C2=O)NC(=O)C(C3=CC=CC=C3)N)C(=O)O)C','Name':'Ampicilline'}, ignore_index=True)#Ampicilline >>> print [str(x) for x in antibiotics.columns] ['Name', 'Smiles'] >>> print antibiotics Name Smiles 0 Penicilline G CC1(C(N2C(S1)C(C2=O)NC(=O)CC3=CC=CC=C3)C(=O)O)C 1 Tetracycline CC1(C2CC3C(C(=O)C(=C(C3(C(=O)C2=C(C4=C1C=CC=C4... 2 Ampicilline CC1(C(N2C(S1)C(C2=O)NC(=O)C(C3=CC=CC=C3)N)C(=O... a new column can be created holding the respective RDKit molecule objects. The fingerprint can be included to accelerate substructure searches on the dataframe. >>> PandasTools.AddMoleculeColumnToFrame(antibiotics,'Smiles','Molecule',includeFingerprints=True) >>> print [str(x) for x in antibiotics.columns] ['Name', 'Smiles', 'Molecule'] A substructure filter can be applied on the dataframe using the RDKit molecule column, because the ">=" operator has been modified to work as a substructure check. Such the antibiotics containing the beta-lactam ring "C1C(=O)NC1" can be obtained by >>> beta_lactam = Chem.MolFromSmiles('C1C(=O)NC1') >>> beta_lactam_antibiotics = antibiotics[antibiotics['Molecule'] >= beta_lactam] >>> print beta_lactam_antibiotics[['Name','Smiles']] Name Smiles 0 Penicilline G CC1(C(N2C(S1)C(C2=O)NC(=O)CC3=CC=CC=C3)C(=O)O)C 2 Ampicilline CC1(C(N2C(S1)C(C2=O)NC(=O)C(C3=CC=CC=C3)N)C(=O... It is also possible to load an SDF file can be load into a dataframe. >>> sdfFile = os.path.join(RDConfig.RDDataDir,'NCI/first_200.props.sdf') >>> frame = PandasTools.LoadSDF(sdfFile,smilesName='SMILES',molColName='Molecule',includeFingerprints=True) >>> frame.info # doctest: +SKIP Int64Index: 200 entries, 0 to 199 Data columns: AMW 200 non-null values CLOGP 200 non-null values CP 200 non-null values CR 200 non-null values DAYLIGHT.FPG 200 non-null values DAYLIGHT_CLOGP 200 non-null values FP 200 non-null values ID 200 non-null values ISM 200 non-null values LIPINSKI_VIOLATIONS 200 non-null values NUM_HACCEPTORS 200 non-null values NUM_HDONORS 200 non-null values NUM_HETEROATOMS 200 non-null values NUM_LIPINSKIHACCEPTORS 200 non-null values NUM_LIPINSKIHDONORS 200 non-null values NUM_RINGS 200 non-null values NUM_ROTATABLEBONDS 200 non-null values P1 30 non-null values SMILES 200 non-null values Molecule 200 non-null values dtypes: object(20)> In order to support rendering the molecules as images in the HTML export of the dataframe, the __str__ method is monkey-patched to return a base64 encoded PNG: >>> molX = Chem.MolFromSmiles('Fc1cNc2ccccc12') >>> print molX # doctest: +SKIP Mol This can be reverted using the ChangeMoleculeRendering method >>> ChangeMoleculeRendering(renderer='String') >>> print molX # doctest: +SKIP >>> ChangeMoleculeRendering(renderer='PNG') >>> print molX # doctest: +SKIP Mol i(tChem(tDraw(t b64encode(tStringIONt.it0ii s display.widthiʚ;sdisplay.heightsdisplay.max_colwidthc"Kstjjj|dddddddddtdtddd dd dd dd td ddddtdddt}|j|jj }|S(sf Patched default escaping of HTML control characters to allow molecule image rendering dataframes tbuftcolumnst col_spacetcolSpacetheadertindextna_reptNaNt formatterst float_formattsparsifyt index_namestjustifyt force_unicodet bold_rowstclassestescapeN( tpdtcoretformattDataFrameFormattertNonetTruetFalsetto_htmlRtgetvalue(tselftkwargst formatterthtml((sX/public2/ftp/pub/ccp4/6.5/ccp4-6.5/lib/python2.7/site-packages/rdkit/Chem/PandasTools.pytpatchPandasHTMLreprhs 6! icCs8|| }tjt||_tjt||_|S(s;Ensure inheritance of patched to_html in "head" subframe (ttypest MethodTypeR$RtpatchPandasHeadMethodthead(R tntdf((sX/public2/ftp/pub/ccp4/6.5/ccp4-6.5/lib/python2.7/site-packages/rdkit/Chem/PandasTools.pyR'ss cCsddl}t}|j|ddt|j}|jdt|d|jdt|dt|d|jd kr|jd t|dn|S( s4displayhook function for PIL Images, rendered as PNGiNRtPNGsdisplay.max_columnsisdisplay.max_rowsidsdisplay.max_colwidth(tpandasRtsaveRRt set_optiontlent get_option(txRtsiots((sX/public2/ftp/pub/ccp4/6.5/ccp4-6.5/lib/python2.7/site-packages/rdkit/Chem/PandasTools.pyt _get_image{s  (t DataStructs(t pyAvalonToolscCstj|d|dtjS(NtisQuerytbitFlags(R6t GetAvalonFPt avalonSSSBits(R1ty((sX/public2/ftp/pub/ccp4/6.5/ccp4-6.5/lib/python2.7/site-packages/rdkit/Chem/PandasTools.pytscCstj|ddS(NtfpSizei(RtPatternFingerprint(R1R;((sX/public2/ftp/pub/ccp4/6.5/ccp4-6.5/lib/python2.7/site-packages/rdkit/Chem/PandasTools.pyR<scCs{|dks|dkrtSt|drnt|dsOt|t|_ntj|j|jsntSn|j|S(s!Allows for substructure check using the >= operator (X has substructure Y -> X >= Y) by monkey-patching the __ge__ function This has the effect that the pandas/numpy rowfilter can be used for substructure filtering (filtered = dframe[dframe['RDKitColumn'] >= SubstructureMolecule]) t _substructfpN( RRthasattrt_fingerprinterRR?R5tAllProbeBitsMatchtHasSubstructMatch(R1R;((sX/public2/ftp/pub/ccp4/6.5/ccp4-6.5/lib/python2.7/site-packages/rdkit/Chem/PandasTools.pyt_molgescCsdttj|S(s4returns the molecules as base64 encoded PNG image s/Mol(R4Rt MolToImage(R1trenderer((sX/public2/ftp/pub/ccp4/6.5/ccp4-6.5/lib/python2.7/site-packages/rdkit/Chem/PandasTools.pytPrintAsBase64PNGStringscCst|jS(N(tstrt__repr__(R1((sX/public2/ftp/pub/ccp4/6.5/ccp4-6.5/lib/python2.7/site-packages/rdkit/Chem/PandasTools.pytPrintDefaultMolRepscCs4tj|}|dk r0t|t|_n|S(sfPrecomputes fingerprints and stores results in molecule objects to accelerate substructure matching N(Rt MolFromSmilesRRARR?(tsmitm((sX/public2/ftp/pub/ccp4/6.5/ccp4-6.5/lib/python2.7/site-packages/rdkit/Chem/PandasTools.pyt_MolPlusFingerprintFromSmiless cCs1|rttjjj_nttjjj_dS(sAChanges the default dataframe rendering to not escape HTML characters, thus allowing rendered images in all dataframes. IMPORTANT: THIS IS A GLOBAL CHANGE THAT WILL AFFECT TO COMPLETE PYTHON SESSION. If you want to change the rendering only for a single dataframe use the "ChangeMoleculeRendering" method instead. N(R$RRtframet DataFrameRtdefPandasRendering(timages((sX/public2/ftp/pub/ccp4/6.5/ccp4-6.5/lib/python2.7/site-packages/rdkit/Chem/PandasTools.pytRenderImagesInAllDataFramesstSmilestROMolcs^|s+|jfddd||st_NameR RVRYRZN(RttypeRHtopent enumerateRtForwardSDMolSuppliertdictt GetPropNamestHasPropRct MolToSmilesRRPtappendtcloseR[( tfilenamet smilesNametidNamet molColNameRZR*tftitrow((RfsX/public2/ftp/pub/ccp4/6.5/ccp4-6.5/lib/python2.7/site-packages/rdkit/Chem/PandasTools.pytLoadSDFs$" "   (t SaltRemovercs&|jfddd|=0.10 not found, skipping testst optionflags(Et__doc__trdkitRt rdkit.ChemRtbase64RRR%R,RtversiontsplittvtintRRtconfigt_registered_optionsR.RORPRRQt ExceptionteR$R'R4R5t rdkit.AvalonR6RAt ImportErrorRDR]t__ge__RGRJR^RNRRSRR[R_RyRzR{R}RRtrdkit.Chem.ScaffoldsRRoRRRRt__name__tsyststderrtdoctestttestmodtELLIPSIStNORMALIZE_WHITESPACEtfailedttriedtexit(((sX/public2/ftp/pub/ccp4/6.5/ccp4-6.5/lib/python2.7/site-packages/rdkit/Chem/PandasTools.pytNsx  &                       & "