o ߥi'@sdZddlZddlZddlZddlmZmZmZddlZ ddl m Z dZ idgdgdgd gd gd gd gdgd gdgdgd gdgdgdgdgd gdgdgdgd gdgdgdgdgdgdgdgdgdgdgdgdgdgd gdgdgdgdgd gd!gd"gdgdgd#gdgd gd$gd%gd&gd'ggdgdggdgdggdgd(Z gd)gd*gd+gd+gd,gd-gd-gd)gd+gd+gd+gd*gd-gd+gd+gd,gd,gd+gd+gd,gZgd)gd)gd)gd.gd)gd)gd/gd)gd)gd)gd)gd)gd)gd.gd)gd)gd)gd)gd.gd)gd)gZidgd0gd1gd2gd3gd4gdgd5gd1gd6gd7gd8gd9gd:gd;gd<gd=gd>g d gd?gd1gd2gd@gdAgdBgdCgdDgdgdEgd1gdFgdGgdHgdIgdJgdKgdgdLgd1gdMgdNgdOgdPgdgdQgd1gdRgdSgdHgdTgdUgdVgdWg dgdXgd1gd2gdYgd8gdZgd[gd\gd]g dgd^gd1gd_gdHgdgd`gd1gdagdbgdcgddgdegdfgdggdhg dgdigd1gdjgdkgdlgdmgdngdogdgdpgd1gd6gdqgdrgdsgdtgdugdgdvgd1gdRgdwgdHgdxgdygdzgd{g dgd|gd1gdagd}gdOgd~gdgdgd"gdgd1gdgdgdHgdgdgdgdgdgdg d#gdgd1gdgdgdgdgdgd$gdgd1gd6gdgdHgdgd&gdgd1gdgdgd8gdgdggdgd1gdgdgd4gdgdgdgdgdgdgdgdgdggdgd1gdgdgdlgdgdgdgdgdgdgdg gdgd1gdjgdgdlgdgdgd(Zidgddgddgdd gddgddgddgddgddgddgddgddgddgdd"gdd#gdd$gdd&gd¢gdâgdĢgdŢd(Zddiddiddd̜ddd̜d͜ZdddddҜZedgdԢZedgd֢Zejdd׍ddلZddgZddgZddgZddgZgdZddeeDZe eZ!idgddgdd gddgddgddgddgddgddgddgddgddgddgdd"gdd#gdd$gdd&gdgdgdgdgddZ"gdZ#ddee#DZ$e e#Z%e%Z&e#dgZ'ddee'DZ( d^de)dee)e*fde+de j,fddZ-idddddd d dd dd dd dd ddddddddddddd"dd#dd$dd&ddddZ.dde./DZ0dZ1dde#De1gZ2ddee2DZ3iddd d!d d"d d!d d#dd$d d%dd&dd'd(d)dd*dd+dd,dd-d.d)dd/d d0d1d2d3d"d4d5d)d6d#d7d8 Z4iddd"d d!d d#d d$dd%d d&dd'dd*dd+dd,dd-dd/dd0d d1dd2dd3dd9d:d;dd<d=Z5e#ddd?e8e e6DZ9de j,fd@dAZ:e:Z;dBe*de j,fdCdDZdEde#DZ?e dFdGe?e)dHZ?e @dIde?DZ?eAeBZCe /D]$\ZDZEeeED]\ZFZGeeGD]\ZHZIeCeDeIfJeFeHfqqqeKeCZCdJdKZLe jMd7dLge jNdMZOe jMd7dLge jPdMZQe jMgdNe jPdMZRe jMd7d1ge jNdMZSe jMd7d1ge jPdMZTe jMgdOe jPdMZUe jMgdPe jPdMZVdQdRZWeW S 2d_dTdUZXdVdWZYeY\ZZZ[ZTdXdYZ\e\Z]dZd[Z^e^Z_d\d]Z`e`ZadS(`zConstants used in AlphaFold.N)ListMappingTuple)tree_mapgj@ALAARG)NCACBCG)r r r CD)r r r NE)r r r CZASN)r r r OD1ASPCYS)rr r SGGLN)r r r OE1GLUGLYHIS)r r r ND1ILE)rr r CG1)r r rCD1LEU)r r r rLYS)r r r CE)r r rNZMET)r r r SD)r r r"rPHEPROSER)rr r OGTHR)rr r OG1)TRPTYRVAL)r,r,r,)?r-r-r-)r-r-r,r,)r-r,r,r,)r-r-r-r,)r,r-r,r,)r,r,r-r,)rr)+?r,)r r)r,r,r,)Cr)"~j?r2)r r)!rhS㥛gHzG)O)Mb?ˡE?r,)rr)S㥛n?r2)r0r)ffffff?r2r2)r r)r9L7A`gMbX)r5r6)x&1?r8r,)r )gʡE?= ףp=?r2)r )g?5^I ?g9v?r,)r )gsh|??ʡE?r2)NH1)g|?5^?g rh@r,)NH2rD)g9v@gK?r2)rrD)gK7A?grh|?r2)rr)x&rBr,)r r)gˡEgv/g333333)r5r6)?r8r,)r r>)g㥛 ?Mb?r,)ND2r@)`"?g5^I gMbP?)rr@)gK7A?gl?r,)rr)r.r:r2)r0r);On?r,r2)r r)Er<S)r5r6)r=r8r2)r r>)rJg|?5^?r2)rr@)gQ?~jt?r,)OD2r@)gl?gV-~jth)rr)NbX9r:r2)r0r)Mb?r,r,)r r)gS㥛#~j1Zd)r5r6)rGr8r2)rr>)gK?g r?r,)rr)rL-?r2)r0r)r1r,r,)r r)r.g!rhZd;O)r r>)Gz?g}?5^I?r,)r r@)gbX9?rHr2)NE2rA)rJg/$MbP)rrA)g}?5^I?g(\?r,)rr)gL7A`rUr,)r r)rLgˡErV)r r>)rWgʡE?r,)r r@)333333?g'1Z?r,)rrA)Cl?gQ?r2)OE2rA)g +?gDlrY)rr)gMg1Zd?r,)r0r)gʡE?r2r2)rr)gw/(\?r,)r0r)r;r,r,)r r)r.r<rM)r5r6)rG5^I ?r,)r r>)rZQ?r2)CD2r@)g r?gtVg~jth?)rr@)g+?g(\?r2)CE1r@)g= ףp=@gZd;?gMb`?)rXr@)g)\(@gmݿgMbp?)rr)gZd;O߿+?r2)r0r)rKr2r2)r r)rFK7A` rh)r5r6)r7r8r2)rr>)gJ +?gˡE?r2)CG2r>)HzG?gQgv/)rr@)+?K7A?r,)rr)gp= ףr/r,)r r)rQrSgCl)r5r6)rGr^r2)r r>)gV-?gV-?r,)rr@)g(\?gzG?r2)r`r@)gQ?r4333333?)rr)rLr:r2)r r)r9r<rM)r r>)rgr?r,)r r@)gS?gy&1?r,)rrA)gQ?g-?r,)r rD)g|?5^?gx&1?r,)rr)y&1m?r2)r r)#~jEg\(\)r r>)gV-?rhr2)r"r@)g"~?gQ?r,)rrA)g{Gz?g~jt?r2)rr)~jtr/r,)r0r)rRr,r2)r r)r.rmrT)r r>)r[gx&1?r,)rr@)g㥛 ?Q?r2)r`r@)gd;O?gA`"r,)rar@)g7A`@g^I +?r2)CE2r@)gbX9@K7r2)rr@)g'1Z@rPrY)rr)g&1gV-?r2)r0r)rKr2r,)r r)gFxgZd;Og㥛 )r5r6)gZd;?guV?r,)r r>)g r?gQ?r,)r r@)gI +?gbX9?r,)rr)r3r]r2)r r)rngw/gK7A`)r&r>)gjt?g333333?r,)rr)g%Crkr,)r0r)r1r,r2)r r)gPnrcgq= ףp)rer>)g?g`"gS)r(r>)gh|?5?gS?r,)rr)rjr/r,)r0r)r;r2r,)r r)rlrmrT)r r>)grh|?r_r2)rr@)g|?5^?rNr,)r`r@)gS?g|?5^{Gzt)rpr@)grh|@gV-y&1|)CE3r@)gCl?g= ףp=rs)NE1r@)gQ@gGz?gMbp)CH2r@)gMbX9@gQg9v)CZ2r@)g$C @g㥛 I +)CZ3r@)gq= ףp?g&1 rx)rr)rQr:r,)r0r)rRr2r2)r r)rQrmrd)r r>)r[g&1?r2)rr@)gx?ror2)r`r@)g7A`?gvrY)rar@)gB`"@riMb`)rpr@)gE@rqrP)OHr@)gy&1@rzrr)rr@)gS@rYrP)rr)gV-߿rbr2)r r)guVgq= ףprd)rr>)rfgw/?r2)rer>)guV?rmgsh|??)r0r r rr5) r0r r r r rrr r5rCrE)r0r r r rr5rrO)r0r r r rrIr5r)r0r r rr5r) r0r r r r rr5rr\) r0r r r r rrXr5r)r0r rr5) r0r r r r`rarrrXr5)r0r r rrerrr5)r0r r r rr`rr5) r0r r r r rrr r5)r0r r r rrr5r") r0r r r rr`rarprrr5)r0r r r r rr5)r0r r rr5r&)r0r r rerr5r()r0r r r rr`rprtrwryrvrrur5) r0r r r rr`rarprrr5r{)r0r r rrerr5rrOrr\r`rp)rra)rrr#r*g333333?g?gRQ?g?)r0rr5SBond) atom1_name atom2_namelengthstddev BondAngle)r~r atom3name angle_radr)maxsizec Cstjtjtjtd}t|ddd }|}Wdn1s%wYt| }i}t ||D]1}| dkrBn(| \}}}} | d\} } ||vrYg||<|| t| | t|t| q8g|d<i} t |t ||D]=}| dkrn4| \}}} }| d\} } }|| vrg| |<| | t| | |t| dtjt|dtjqzg| d<d d }i}| D]\}}i}||D] }||||j|j<qg||<|D]}|||j|j}|||j|j}|j}t|jd |jd d |j|jt|}d |}d |j|jt||}d |jd |jt||}d |jd |jt||}t||jd ||jd ||jd } || t|j|j|| qq||| fS) aLoad stereo_chemical_props.txt into a nice structure. Load literature values for bond lengths and bond angles and translate bond angles into the length of the opposite edge of the triangle ("residue_virtual_bonds"). Returns: residue_bonds: Dict that maps resname -> list of Bond tuples. residue_virtual_bonds: Dict that maps resname -> list of Bond tuples. residue_bond_angles: Dict that maps resname -> list of BondAngle tuples. zstereo_chemical_props.txtrtzutf-8)encodingN-UNKgf@cSsdt||gS)zUnique key to lookup bonds.r)joinsorted)r~rrl/home/ubuntu/.local/lib/python3.10/site-packages/modelscope/models/science/unifold/data/residue_constants.py make_bond_keysz1load_stereo_chemical_props..make_bond_keyg?)ospathrdirnameabspath__file__openreaditer splitlinesnextstripsplitappendr}floatrnppiitemsr~rrrsqrtrcossinr)stereo_chemical_props_pathfstereo_chemical_props lines_iter residue_bondslinebondresnamerratom1atom2residue_bond_angles angle_degree stddev_degreeatom3rresidue_virtual_bonds bond_angles bond_cachebbabond1bond2gammadl_outer dl_dgammadl_db1dl_db2rrrload_stereo_chemical_propss        rg$C?g~jt?gy&1?gMb?gJ{/LgPs?g -ܿgU؟?)%rr r0r r5r rrer&r(rr rr`rrIrrOr"rrarprtr rurXrr\rvrCrEr{rrwryr OXTcCi|]\}}||qSrr).0i atom_typerrr =r)rr r0r5r rrrrrrrr)rr r0r5r r r r rrCrErrr)rr r0r5r r rrIrrrrrr)rr r0r5r r rrOrrrrrr)rr r0r5r rrrrrrrrr)rr r0r5r r r rrXrrrrr)rr r0r5r r r rr\rrrrr)rr r0r5rrrrrrrrrr)rr r0r5r r rr`rarXrrrr)rr r0r5r rrerrrrrrr)rr r0r5r r rr`rrrrrr)rr r0r5r r r rr rrrrr)rr r0r5r r r"rrrrrrr)rr r0r5r r rr`rarprrrr)rr r0r5r r r rrrrrrr)rr r0r5r r&rrrrrrrr)rr r0r5r r(rerrrrrrr)rr r0r5r r rr`rurprtrwryrv)rr r0r5r r rr`rarprr{rr)rr r0r5r rrerrrrrrr)rrrrrrrrrrrrrr)r)r*r+r)ARrDr0QEGHILKMFPr|TWYVcCrrrrrrestyperrrrrXcCrrrrrrrrsFsequencemappingmap_unknown_to_xreturncCst|d}tt|tt|kr tdt|tjt ||ftj d}t |D](\}}|rN| rG| rG|||d}n td|||}d|||f<q0|S)aMaps the given sequence into a one-hot encoded matrix. Args: sequence: An amino acid sequence. mapping: A dictionary mapping amino acids to integers. map_unknown_to_x: If True, any amino acid that is not in the mapping will be mapped to the unknown amino acid 'X'. If the mapping doesn't contain amino acid 'X', an error will be thrown. If False, any amino acid not in the mapping will throw an error. Returns: A numpy array of shape (seq_len, num_unique_aas) with one-hot encoding of the sequence. Raises: ValueError: If the mapping doesn't contain values from 0 to num_unique_aas - 1 without any gaps. zQThe mapping must have values from 0 to num_unique_aas-1 without any gaps. Got: %sdtyperz#Invalid character in the sequence: )maxvaluesrsetlistrange ValueErrorrzeroslenint32 enumerateisalphaisupperget)rrr num_entries one_hot_arraa_indexaa_typeaa_idrrrsequence_to_onehots$ rrrrrr0rrrrrrrrrrr|rr)r*r+)rrrcCrrr)rkvrrrrrrcCg|]}t|qSr restype_1to3rrrrr !rcCrrr)rrrrrrr"rBrrr6r>r@rArDJ r5 ) rr|rUrrrrZrrrrr)r r r rr ccs|] }tt|VqdSN)restypes_with_x_and_gapindexID_TO_HHBLITS_AA)rrrrr ds  rcCs\tjtdtgtjd}ttD]\}}t|}t|}|D] }t |}d|||f<qq|S)z3Returns [num_res_types, num_atom_types] mask array.rr) rr restype_num atom_type_numrrrestypesr residue_atoms atom_order)maskrrestype_letter restype_name atom_names atom_namerrrr_make_standard_atom_maskisr atom_indexc si}g}tD]\}}fdd|D}|dgdt||||<qtD]}t|}tt||}| |q(| t dtgtj |dd}t |gd}|S)z:Define chi-angle rigid groups via one-hot representations.csg|] }t|qSr) atom_typesr)rsr rrrsz"chi_angle_atom..r>raxis)rrr) chi_angles_atomsrextendrrrreyerrrstack transpose) r chi_angles_indexone_hotsrrindicesrres3one_hotrr#rchi_angle_atom{s  r1cCsg|]}tt|qSr)r'rrrrrrcCst|Srr)nrrrsr5) leaf_typecCs&g|]}|gdgdt|qS)rrrrr>)r)r chi_atomsrrrrscCsr|tj|}|t|||}|tj|}t||}t||||g}tj|gdggdd}|S)zBCreate a rigid 4x4 transformation matrix from two axes and transl.)r,r,r,r-rr%)rlinalgnormdotcrossr*r+ concatenate)exey translation ex_normalized ey_normalizedeznormmrrr_make_rigid_transformation_4x4s  rE%r)r rFr6)r rr6)r rr>r>c s\ttD]G\}}t|}t|D]:\}}}t|}|t||f<dt||f<|t||ddf<t| |}|t ||f<dt ||f<|t ||ddf<qqttD]\}}t|}ddt|Dt dt|dddddf<t dt|dddddf<tddt gd dd }|t|d ddddf<td dddd d }|t|d ddddf<t|drt|d} fdd| D} t| d | d| d| d| d d }|t|dddddf<tddD]/} t|| r*t|| d } | } t| t gd| d }|t|d| ddddf<qqPdS)zFill the arrays above.rNcSsi|] \}}}|t|qSr)rarray)rname_posrrrrs z/_make_rigid_group_constants..r>rrr )r-r,r,)r>r?r@rr0r6cg|]}|qSrrrrHatom_positionsrrrsz/_make_rigid_group_constants..)gr,r,)rrrrigid_group_atom_positionsrrestype_atom37_to_rigid_grouprestype_atom37_mask$restype_atom37_rigid_group_positionsrestype_name_to_atom14_namesrrestype_atom14_to_rigid_grouprestype_atom14_mask$restype_atom14_rigid_group_positionsrr)!restype_rigid_group_default_framerErGchi_angles_maskr'r)rrratomname group_idx atom_positionatomtype atom14idxmatbase_atom_namesbase_atom_positionschi_idxaxis_end_atom_nameaxis_end_atom_positionrrMr_make_rigid_group_constantss         rd?cCstgdtj}tgdtj}tgdtj}t\}}}ttD]\}} t| } t| } t| D]G\} } | scompute upper and lower bounds for bonds to assess violations.)r rrrg _B) lower_bound upper_boundr)rrfloat32rrrrrSvan_der_waals_radiusrr~rrr)overlap_tolerancebond_length_tolerance_factorrestype_atom14_bond_lower_boundrestype_atom14_bond_upper_boundrestype_atom14_bond_stddevrrrIrrr atom_list atom1_idxr~ atom1_radius atom2_idxr atom2_radiuslowerupperrrrrmake_atom14_dists_boundss       rvcsg}g}g}tD]1}tt|}|dd|Dddt|D|fddtD|dd|Dq|dgd|dgd |d gdtj|tjd }tj|tjd }tj|tj d }|||fS) NcSsg|] }|r t|ndqSrr3rLrrrrUsz5_make_atom14_and_atom37_constants..cSrrr)rrrHrrrrWrz5_make_atom14_and_atom37_constants..cs g|] }|vr |ndqSrwrrLatom_name_to_idx14rrrXscSsg|]}|rdndqS)r-r,rrLrrrr]srrrFr,r) rrSrrrr!rrGrrh)restype_atom14_to_atom37restype_atom37_to_atom14rUrrrrxr!_make_atom14_and_atom37_constantsNs4     r|c sddtD}|dg7}dd|DtD]@\}}td}|D].\}}t||}t||}|||<|||<td}t|D] \} } d|| | f<qGq$||<qt fd d|D} | S) NcSrrrrresrrrrxrz+_make_renaming_matrices..rcSsi|]}|tdqS)r)rr)r}rrrr|r2z+_make_renaming_matrices..r)rrr-crKrr)rr all_matricesrrrr) rresidue_atom_renaming_swapsrrarangerSrrrr*) restype_3rswapcorrespondencessource_atom_swaptarget_atom_swap source_index target_indexrenaming_matrixrcorrespondencerenaming_matricesrrr_make_renaming_matricesus.    rcCsrtd}tD]-\}}|D]$\}}tt|}t||}t||}d|||f<d|||f<qq |S)N)r rr)rrrr restype_order restype_3to1rSr)restype_atom14_is_ambiguousrr atom_name1 atom_name2r atom_idx1 atom_idx2rrr_make_atom14_is_ambiguouss   rcCsg}tD]2}t|}t|}g}|D] }|dd|Dqtdt|D] }|gdq'||q|gdgd|S)ajReturns atom indices needed to compute chi angles for all residue types. Returns: A tensor of shape [residue_types=21, chis=4, atoms=4]. The residue types are in the order specified in restypes + unknown residue type at the end. For chi angles which are not defined on the residue, the positions indices are by default set to 0. cSrrr3)ratomrrrrrz(get_chi_atom_indices..r>r7)rrr'rrr)chi_atom_indices residue_nameresidue_chi_angles atom_indices chi_anglerIrrrget_chi_atom_indicess  r)F)rer)b__doc__ collections functoolsrtypingrrrnumpyr unicore.utilsrca_car'rXchi_pi_periodicrOrrri namedtupler}r lru_cacherbetween_res_bond_length_c_n"between_res_bond_length_stddev_c_nbetween_res_cos_angles_c_n_cabetween_res_cos_angles_ca_c_nr!rrrrrSrrrunk_restype_indexrestypes_with_xrestype_order_with_xstrintboolndarrayrrrr unk_restyperesnamesresname_to_idxHHBLITS_AA_TO_IDrrtupler MAP_HHBLITS_AATYPE_TO_OUR_AATYPErSTANDARD_ATOM_MASKr1chi_atom_1_one_hotchi_atom_2_one_hotchi_angles_atom_indicesrG defaultdictrchi_groups_for_atomres_namechi_angle_atoms_for_res chi_group_i chi_groupatom_irrdictrErint_rPrhrQrRrTrUrVrWrdrvr|rzr{rrrrrrrrrrs*       %& ' ()1$)1<GMYcmx",U     /  ^',.024DEFHq  ,                           N 8$