RDKit 記述子が 400 に増えたさかい、他と比較してみてん。

RDKit のバージョンが 2020.09.1 になって、計算できる記述子の数が増えたさかい、他と比較してみたで。え？ 今年は 2021 年やって？ ほんまやな。この記事書いてから気付いたわ。ちょっと古い話かも知れへんけど、堪忍やで。

RDKit installer

RDKit installer は、 miniconda を使って RDKit を Google Colaboratory 上にインストールできるパッケージやで。ついでに色んな補助機能も付けてん。偉いやろ。褒めて褒めて。まずは、こんな感じで pip install しはってな。

%%time 
!pip install git+https://github.com/maskot1977/rdkit_installer.git

Collecting git+https://github.com/maskot1977/rdkit_installer.git
  Cloning https://github.com/maskot1977/rdkit_installer.git to /tmp/pip-req-build-geoedzyq
  Running command git clone -q https://github.com/maskot1977/rdkit_installer.git /tmp/pip-req-build-geoedzyq
Building wheels for collected packages: rdkit-installer
  Building wheel for rdkit-installer (setup.py) ... [?25l[?25hdone
  Created wheel for rdkit-installer: filename=rdkit_installer-0.2.0-py3-none-any.whl size=5686 sha256=671153f44f25467ad8d1d23028a8595af9e0ad744c806d548044d2b9573e276a
  Stored in directory: /tmp/pip-ephem-wheel-cache-jhs5sxr1/wheels/e6/72/a5/218f5f909a3a87c1ec1ccec03ac61298947fb5f1efa517eefa
Successfully built rdkit-installer
Installing collected packages: rdkit-installer
Successfully installed rdkit-installer-0.2.0
CPU times: user 41.7 ms, sys: 19.4 ms, total: 61.1 ms
Wall time: 5.46 s

Anaconda は有償化されたねんけど miniconda 使わはる分にはOKでっせ っちゅう記事（参考文献）があったで。知らんけど。

%%time 
from rdkit_installer import install
install.from_miniconda(rdkit_version="2020.09.1")

add /root/miniconda/lib/python3.7/site-packages to PYTHONPATH
python version: 3.7.12
fetching installer from https://repo.continuum.io/miniconda/Miniconda3-4.7.12-Linux-x86_64.sh
done
installing miniconda to /root/miniconda
done
installing rdkit
done
rdkit-2020.09.1 installation finished!


CPU times: user 728 ms, sys: 303 ms, total: 1.03 s
Wall time: 53.6 s

RDKit Descriptor

RDKit 2020.09.1 のバージョンから４００個の記述子が計算できるようになったねんけど、古いバージョンの２０８個の記述子だけ計算したい思わはんねんたら、こうしはったらええで。

化合物のデータは見せれへんけど、df は pandas の DataFrame 形式のデータで、その 'Open Babel SMILES' カラムに SMILES文字列が入ってる思わはって。

%%time 
from rdkit_installer.descriptors import calc_208descriptors

rdkit_208descriptors_df = calc_208descriptors(df['Open Babel SMILES'])
display(rdkit_208descriptors_df)

	MaxEStateIndex	MinEStateIndex	MaxAbsEStateIndex	MinAbsEStateIndex	qed	MolWt	HeavyAtomMolWt	ExactMolWt	NumValenceElectrons	NumRadicalElectrons	MaxPartialCharge	MinPartialCharge	MaxAbsPartialCharge	MinAbsPartialCharge	FpDensityMorgan1	FpDensityMorgan2	FpDensityMorgan3	BCUT2D_MWHI	BCUT2D_MWLOW	BCUT2D_CHGHI	BCUT2D_CHGLO	BCUT2D_LOGPHI	BCUT2D_LOGPLOW	BCUT2D_MRHI	BCUT2D_MRLOW	BalabanJ	BertzCT	Chi0	Chi0n	Chi0v	Chi1	Chi1n	Chi1v	Chi2n	Chi2v	Chi3n	Chi3v	Chi4n	Chi4v	HallKierAlpha	...	fr_hdrzine	fr_hdrzone	fr_imidazole	fr_imide	fr_isocyan	fr_isothiocyan	fr_ketone	fr_ketone_Topliss	fr_lactam	fr_lactone	fr_methoxy	fr_morpholine	fr_nitrile	fr_nitro	fr_nitro_arom	fr_nitro_arom_nonortho	fr_nitroso	fr_oxazole	fr_oxime	fr_para_hydroxylation	fr_phenol	fr_phenol_noOrthoHbond	fr_phos_acid	fr_phos_ester	fr_piperdine	fr_piperzine	fr_priamide	fr_prisulfonamd	fr_pyridine	fr_quatN	fr_sulfide	fr_sulfonamd	fr_sulfone	fr_term_acetylene	fr_tetrazole	fr_thiazole	fr_thiocyan	fr_thiophene	fr_unbrch_alkane	fr_urea
0	12.203893	-0.011059	12.203893	0.011059	0.747310	236.359	212.167	236.188863	96	0	0.232872	-0.315939	0.315939	0.232872	1.352941	2.000000	2.588235	16.153539	10.090561	2.217843	-2.244821	2.189928	-2.340170	5.823874	-0.131085	2.692946	342.131373	12.999636	11.656047	11.656047	7.913591	6.243605	6.243605	5.131783	5.131783	3.017810	3.017810	2.172367	2.172367	-1.09	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
1	8.990151	0.150741	8.990151	0.150741	0.649670	150.221	136.109	150.104465	60	0	0.068649	-0.391660	0.391660	0.068649	1.090909	1.545455	1.909091	16.250348	10.006804	2.001032	-2.068853	2.150011	-1.932082	5.362305	0.279980	3.247713	240.570442	8.430721	7.309021	7.309021	5.147066	3.824482	3.824482	3.035385	3.035385	1.998483	1.998483	1.459315	1.459315	-0.82	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
2	9.393241	-0.207824	9.393241	0.008704	0.608365	168.192	156.096	168.078644	66	0	0.126112	-0.507200	0.507200	0.126112	1.083333	1.500000	1.833333	16.273518	10.058331	2.056393	-2.067148	2.237259	-1.920923	5.421688	0.263338	3.219755	256.097291	9.137828	6.910555	6.910555	5.685071	3.717870	3.717870	2.748003	2.748003	1.752819	1.752819	1.175315	1.175315	-1.06	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
3	9.119284	0.353611	9.119284	0.353611	0.652274	150.221	136.109	150.104465	60	0	0.115360	-0.507956	0.507956	0.115360	1.272727	1.909091	2.363636	16.254670	9.980552	2.018479	-2.082124	2.216557	-1.897398	5.349107	0.474188	3.092720	251.049732	8.430721	7.256615	7.256615	5.109061	3.905016	3.905016	3.195591	3.195591	1.875392	1.875392	1.307668	1.307668	-0.98	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
4	2.401620	0.311389	2.401620	0.311389	0.485103	204.357	180.165	204.187801	84	0	-0.014017	-0.084980	0.084980	0.014017	1.066667	1.800000	2.533333	14.164831	9.873635	2.112923	-2.177669	2.253025	-2.050045	5.115296	0.485569	2.790626	287.160171	11.311555	10.637828	10.637828	6.994800	5.984764	5.984764	5.438827	5.438827	3.016960	3.016960	1.800810	1.800810	-0.78	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
628	8.990463	-0.446481	8.990463	0.057870	0.330838	156.137	148.073	156.042259	60	0	0.163917	-0.504157	0.504157	0.163917	0.909091	1.272727	1.636364	16.319815	10.150606	2.112174	-2.046124	2.403314	-1.779929	5.573520	0.368156	3.455546	266.732315	8.593858	5.866205	5.866205	5.036581	2.971777	2.971777	2.237032	2.237032	1.469088	1.469088	0.762583	0.762583	-1.58	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	4	4	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
629	10.577593	-0.546019	10.577593	0.141481	0.571034	142.110	136.062	142.026609	54	0	0.226354	-0.501988	0.501988	0.226354	1.600000	2.300000	2.800000	16.364841	10.370775	1.977536	-1.919971	2.014312	-1.988382	5.143583	0.240965	3.172371	275.321460	7.560478	5.072731	5.072731	4.736382	2.653364	2.653364	1.753958	1.753958	1.022301	1.022301	0.526670	0.526670	-1.29	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
630	4.165093	1.049167	4.165093	1.049167	0.605928	146.193	136.113	146.084398	56	0	0.093137	-0.344588	0.344588	0.093137	1.181818	1.909091	2.727273	14.858487	10.106251	1.972654	-1.967400	2.143397	-1.794492	5.757300	1.343803	2.857747	351.362105	7.844935	6.679264	6.679264	5.270857	3.675181	3.675181	2.856153	2.856153	2.035718	2.035718	1.194726	1.194726	-1.31	...	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
631	9.567963	0.398056	9.567963	0.398056	0.652274	150.221	136.109	150.104465	60	0	0.121428	-0.507385	0.507385	0.121428	1.272727	1.909091	2.363636	16.256039	9.998067	2.053190	-2.085892	2.271210	-1.884369	5.403987	0.460434	3.165083	251.049732	8.430721	7.256615	7.256615	5.125898	3.910999	3.910999	3.167445	3.167445	1.890976	1.890976	1.147545	1.147545	-0.98	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
632	10.518632	-0.384815	10.518632	0.156667	0.392541	166.180	156.100	166.074228	64	0	0.274814	-0.398706	0.398706	0.274814	1.333333	1.750000	2.083333	16.628418	10.090020	2.134719	-2.013243	2.243835	-1.965347	5.555281	-0.385558	3.379814	310.824900	9.300965	6.995761	6.995761	5.519745	3.532131	3.532131	2.760417	2.760417	1.695964	1.695964	1.277373	1.277373	-1.58	...	0	0	0	0	0	0	0	0	0	0	0	0	0	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0

633 rows × 208 columns

CPU times: user 5.49 s, sys: 265 ms, total: 5.75 s
Wall time: 5.79 s

まぁざっとこんな感じや。ついでやし計算時間も表示するようにしといたで。

そんで、４００個の記述子を計算するコードはこっちや。

%%time 
from rdkit_installer.descriptors import calc_descriptors

rdkit_descriptors_df = calc_descriptors(df['Open Babel SMILES'])
display(rdkit_descriptors_df)

	AUTOCORR2D_1	AUTOCORR2D_10	AUTOCORR2D_100	AUTOCORR2D_101	AUTOCORR2D_102	AUTOCORR2D_103	AUTOCORR2D_104	AUTOCORR2D_105	AUTOCORR2D_106	AUTOCORR2D_107	AUTOCORR2D_108	AUTOCORR2D_109	AUTOCORR2D_11	AUTOCORR2D_110	AUTOCORR2D_111	AUTOCORR2D_112	AUTOCORR2D_113	AUTOCORR2D_114	AUTOCORR2D_115	AUTOCORR2D_116	AUTOCORR2D_117	AUTOCORR2D_118	AUTOCORR2D_119	AUTOCORR2D_12	AUTOCORR2D_120	AUTOCORR2D_121	AUTOCORR2D_122	AUTOCORR2D_123	AUTOCORR2D_124	AUTOCORR2D_125	AUTOCORR2D_126	AUTOCORR2D_127	AUTOCORR2D_128	AUTOCORR2D_129	AUTOCORR2D_13	AUTOCORR2D_130	AUTOCORR2D_131	AUTOCORR2D_132	AUTOCORR2D_133	AUTOCORR2D_134	...	fr_hdrzone	fr_imidazole	fr_imide	fr_isocyan	fr_isothiocyan	fr_ketone	fr_ketone_Topliss	fr_lactam	fr_lactone	fr_methoxy	fr_morpholine	fr_nitrile	fr_nitro	fr_nitro_arom	fr_nitro_arom_nonortho	fr_nitroso	fr_oxazole	fr_oxime	fr_para_hydroxylation	fr_phenol	fr_phenol_noOrthoHbond	fr_phos_acid	fr_phos_ester	fr_piperdine	fr_piperzine	fr_priamide	fr_prisulfonamd	fr_pyridine	fr_quatN	fr_sulfide	fr_sulfonamd	fr_sulfone	fr_term_acetylene	fr_tetrazole	fr_thiazole	fr_thiocyan	fr_thiophene	fr_unbrch_alkane	fr_urea	qed
0	2.962	3.093	-0.045	-0.266	0.351	-0.275	0.073	-0.268	0.111	-0.212	0.096	-0.304	3.205	0.319	-0.295	0.05	-0.182	0.134	-0.208	-0.051	-0.263	0.349	-0.273	3.100	0.074	-0.238	0.129	-0.215	0.034	-0.294	0.349	-0.293	0.061	-0.307	2.600	0.042	-0.187	0.230	-0.298	0.185	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.747310
1	2.512	2.755	-0.300	-0.450	-1.000	0.000	0.000	-0.000	0.027	-0.047	-0.300	-0.450	2.736	-1.000	0.000	0.00	0.000	0.027	-0.047	-0.300	-0.450	-1.000	0.000	2.385	0.000	-0.000	0.027	-0.047	-0.300	-0.450	-1.000	0.000	0.000	0.000	0.999	0.027	-0.047	-0.300	-0.450	-1.000	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.649670
2	2.639	2.764	0.231	-1.000	3.000	0.000	0.000	-0.000	-0.000	-0.294	0.231	-1.000	2.755	3.000	0.000	0.00	0.000	0.000	-0.294	0.231	-1.000	3.000	0.000	2.451	0.000	-0.000	-0.000	-0.294	0.231	-1.000	3.000	0.000	0.000	0.000	1.792	0.000	-0.294	0.231	-1.000	3.000	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.608365
3	2.512	2.736	-0.120	-0.267	-1.000	0.000	0.000	-0.000	-0.047	-0.057	-0.120	-0.267	2.669	-1.000	0.000	0.00	0.000	-0.047	-0.057	-0.120	-0.267	-1.000	0.000	2.345	0.000	-0.000	-0.047	-0.057	-0.120	-0.267	-1.000	0.000	0.000	0.000	1.312	-0.047	-0.057	-0.120	-0.267	-1.000	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.652274
4	2.773	3.045	1.000	1.000	1.000	0.000	0.000	1.000	1.000	1.000	1.000	1.000	2.996	1.000	0.000	0.00	1.000	1.000	1.000	1.000	1.000	1.000	0.000	2.996	0.000	1.000	1.000	1.000	1.000	1.000	1.000	0.000	0.000	1.000	3.091	1.000	1.000	1.000	1.000	1.000	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.485103
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
628	2.590	2.689	-0.214	1.750	0.000	0.000	0.000	-0.000	-0.214	-0.214	-0.214	1.750	2.651	0.000	0.000	0.00	0.000	-0.214	-0.214	-0.214	1.750	0.000	0.000	1.989	0.000	-0.000	-0.214	-0.214	-0.214	1.750	0.000	0.000	0.000	0.000	0.704	-0.214	-0.214	-0.214	1.750	0.000	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	4	4	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.330838
629	2.539	2.461	-0.306	-0.167	1.500	0.000	0.000	-0.167	-0.359	0.181	-0.306	-0.167	2.313	1.500	0.000	0.00	-0.167	-0.359	0.181	-0.306	-0.167	1.500	0.000	1.681	0.000	-0.167	-0.359	0.181	-0.306	-0.167	1.500	0.000	0.000	-0.167	1.079	-0.359	0.181	-0.306	-0.167	1.500	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.571034
630	2.615	2.810	-0.389	-0.389	0.000	0.000	0.000	-0.185	0.186	-0.127	-0.389	-0.389	2.641	0.000	0.000	0.00	-0.185	0.186	-0.127	-0.389	-0.389	0.000	0.000	2.083	0.000	-0.185	0.186	-0.127	-0.389	-0.389	0.000	0.000	0.000	-0.185	1.508	0.186	-0.127	-0.389	-0.389	0.000	...	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.605928
631	2.512	2.736	-0.230	0.100	0.000	0.000	0.000	-0.000	-0.047	-0.193	-0.230	0.100	2.699	0.000	0.000	0.00	0.000	-0.047	-0.193	-0.230	0.100	0.000	0.000	2.317	0.000	-0.000	-0.047	-0.193	-0.230	0.100	0.000	0.000	0.000	0.000	1.609	-0.047	-0.193	-0.230	0.100	0.000	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.652274
632	2.670	2.771	-0.715	-0.713	1.290	0.000	0.000	0.510	0.103	-0.364	-0.725	-0.233	2.735	1.950	0.000	0.00	0.423	0.231	-0.273	-0.712	-0.725	1.249	0.000	2.546	0.000	0.495	0.158	-0.333	-0.736	-0.480	1.755	0.000	0.000	0.423	1.100	-0.019	-0.395	-0.611	0.507	1.809	...	0	0	0	0	0	0	0	0	0	0	0	0	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0.392541

633 rows × 400 columns

CPU times: user 6.03 s, sys: 332 ms, total: 6.37 s
Wall time: 6.36 s

なんや、案外、時間かからへんな。せやけどこのカラム名、何やねん。意味わからん特徴量が増えたな。解釈性あきらめたんかな。知らんけど。

各種フィンガープリント

ついでやし、各種フィンガープリントも計算してまうで。

%%time 
from rdkit_installer.fingerprints import Fingerprinter

fingerprinter = Fingerprinter()
fp_dfs = []
for name in fingerprinter.names:
    fp_df = pd.DataFrame([vec for vec in fingerprinter.transform(df['Open Babel SMILES'], fp_type=name)])
    if name in fingerprinter.all_bit_info_keys.keys():
        fp_df.columns=fingerprinter.all_bit_info_keys[name]

    print(name, fp_df.shape)
    fp_dfs.append(fp_df)

MACCSkeys (633, 167)
Avalon (633, 512)
Morgan2(1024bits) (633, 1024)
Morgan2F(1024bits) (633, 1024)
Morgan4(2048bits) (633, 2048)
Morgan4F(2048bits) (633, 2048)
ECFP2 (633, 693)
FCFP2 (633, 207)
ECFP4 (633, 3128)
FCFP4 (633, 1450)
ECFP6 (633, 6508)
FCFP6 (633, 4110)
CPU times: user 12.5 s, sys: 401 ms, total: 12.9 s
Wall time: 12.7 s

できあがったフィンガープリントはこんな感じや。ぎょうさんあっから見んのもえらいけど、行くで。１か０しかあらへん「bit vector」と、それ以外の自然数もある「count vector」があるさかい、違いをよお見とかはるんやで。

for fp_df, name in zip(fp_dfs, fingerprinter.names):
    print("#", name)
    display(fp_df)
    print()

# MACCSkeys

	0	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25	26	27	28	29	30	31	32	33	34	35	36	37	38	39	...	127	128	129	130	131	132	133	134	135	136	137	138	139	140	141	142	143	144	145	146	147	148	149	150	151	152	153	154	155	156	157	158	159	160	161	162	163	164	165	166
0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	1	0	0	0	0	0	0	0	0	0	1	1	0	0	1	1	1	0	0	0	1	1	1	1	0	0	1	1	1	1	0	1	0	1	1	0	0	1	1	0
1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	1	0	1	0	0	1	0	0	0	0	1	1	0	0	1	0	1	0	1	0	0	1	0	1	1	1	1	0
2	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	1	1	0	0	1	0	0	0	0	0	0	1	1	0	0	0	1	1	0	1	0	0	0	1	0	1	1	0	1	0	1	0	1	1	0	1	1	1	1	0
3	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	1	0	0	0	0	0	0	0	0	0	0	0	1	0	1	0	1	1	0	0	0	0	1	1	0	1	0	0	0	0	1	0	0	1	0	1	1	1	1	0
4	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	1	1	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	1	0	1	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	1	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
628	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	1	0	0	0	1	0	0	0	0	0	0	0	1	1	0	0	1	1	0	1	0	0	0	1	0	1	0	0	0	0	1	0	1	1	0	1	1	1	1	0
629	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	1	0	0	0	1	1	0	0	0	0	1	0	1	1	0	0	1	1	0	1	0	0	0	1	0	1	1	1	1	0	1	0	1	0	0	1	1	1	1	0
630	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	1	0	1	0	0	0	0	1	1	1	0	0	0	0	1	0	0	0	1	1	1	1	0	1	0
631	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	1	0	0	0	0	0	0	0	0	0	0	0	1	0	1	0	1	1	0	0	0	0	1	1	0	1	0	0	0	0	1	0	0	1	0	1	1	1	1	0
632	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	1	0	0	1	0	1	0	0	0	0	0	0	1	0	1	0	0	0	1	1	1	1	0	0	0	0	1	0	1	1	1	1	1	1	1	1	0

633 rows × 167 columns

# Avalon

	0	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25	26	27	28	29	30	31	32	33	34	35	36	37	38	39	...	472	473	474	475	476	477	478	479	480	481	482	483	484	485	486	487	488	489	490	491	492	493	494	495	496	497	498	499	500	501	502	503	504	505	506	507	508	509	510	511
0	0	0	0	0	1	1	0	0	0	0	1	0	0	1	0	1	0	0	0	0	0	1	0	1	0	0	0	1	0	1	1	0	0	0	1	0	1	0	0	1	...	1	0	0	0	1	0	0	0	0	0	0	0	0	0	1	0	1	0	1	0	0	0	1	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	1	0
1	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	1	0	0	0	0	0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
2	0	0	0	0	0	0	0	0	0	0	0	0	0	1	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	1	1	0	0	0	0	0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
3	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	...	0	0	0	0	0	0	1	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0
4	0	0	0	0	0	1	0	0	0	0	1	0	0	0	0	0	0	0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	...	0	0	0	0	0	0	1	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	1	0	0	0	0	0	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
628	0	0	0	0	0	0	0	0	0	0	1	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	...	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0
629	0	0	0	1	1	1	0	0	1	0	0	0	0	1	0	0	1	0	0	1	0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	1	0	0	...	0	0	1	1	1	1	1	1	0	0	0	0	0	1	1	0	1	0	1	0	0	1	0	0	0	0	0	1	0	0	1	0	0	0	0	0	0	0	0	0
630	1	0	0	0	0	0	0	0	0	1	1	0	0	1	0	1	1	1	0	0	1	0	0	0	0	0	1	1	1	1	0	0	0	0	1	0	0	0	0	1	...	0	0	0	1	1	0	0	1	0	1	0	0	1	1	0	0	1	0	1	0	1	1	0	0	0	0	0	1	0	0	0	1	1	0	0	1	1	0	1	1
631	0	0	0	0	0	0	0	0	0	0	1	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
632	0	0	0	0	0	0	0	0	0	1	0	0	0	1	0	1	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	1	0	0	0	0	1	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	1	0	0	0	0	0	0

633 rows × 512 columns

# Morgan2(1024bits)

	0	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25	26	27	28	29	30	31	32	33	34	35	36	37	38	39	...	984	985	986	987	988	989	990	991	992	993	994	995	996	997	998	999	1000	1001	1002	1003	1004	1005	1006	1007	1008	1009	1010	1011	1012	1013	1014	1015	1016	1017	1018	1019	1020	1021	1022	1023
0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0
1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0
2	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
3	0	1	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
4	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	1	0	0	1	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
628	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0
629	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
630	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0
631	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
632	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0

633 rows × 1024 columns

# Morgan2F(1024bits)

	0	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25	26	27	28	29	30	31	32	33	34	35	36	37	38	39	...	984	985	986	987	988	989	990	991	992	993	994	995	996	997	998	999	1000	1001	1002	1003	1004	1005	1006	1007	1008	1009	1010	1011	1012	1013	1014	1015	1016	1017	1018	1019	1020	1021	1022	1023
0	1	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0
1	1	0	0	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0
2	1	0	0	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0
3	1	0	0	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0
4	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
628	1	0	0	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	1	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
629	1	0	1	1	1	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	1	0	0	0	0	0	0
630	1	0	0	0	1	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0
631	1	0	0	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
632	1	0	1	0	1	0	0	0	0	0	0	0	0	0	0	1	1	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0

633 rows × 1024 columns

# Morgan4(2048bits)

	0	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25	26	27	28	29	30	31	32	33	34	35	36	37	38	39	...	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022	2023	2024	2025	2026	2027	2028	2029	2030	2031	2032	2033	2034	2035	2036	2037	2038	2039	2040	2041	2042	2043	2044	2045	2046	2047
0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0
2	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
3	0	1	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
4	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
628	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0
629	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
630	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0
631	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
632	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	1	0	0	1	0	0	0	0	0	0	0

633 rows × 2048 columns

# Morgan4F(2048bits)

	0	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25	26	27	28	29	30	31	32	33	34	35	36	37	38	39	...	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022	2023	2024	2025	2026	2027	2028	2029	2030	2031	2032	2033	2034	2035	2036	2037	2038	2039	2040	2041	2042	2043	2044	2045	2046	2047
0	1	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0
1	1	0	0	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0
2	1	0	0	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	...	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0
3	1	0	0	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
4	1	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
628	1	0	0	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	1	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
629	1	0	1	1	1	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	1	0	0	0	0	0	0
630	1	0	0	1	1	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0
631	1	0	0	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
632	1	0	1	0	1	0	0	0	0	0	0	0	0	0	0	1	1	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0

633 rows × 2048 columns

# ECFP2

	2245273601	2583443460	3925172229	3325788165	918112266	4008337421	2088929298	1150554131	2027655192	478283801	148058140	737505311	3327193120	2251845666	3820826669	3163127853	1533767736	1909452861	1446633534	1821515839	2681681991	694069322	74537039	2245384272	1512794193	1445828689	2082340947	994494548	469172305	3808467027	1369337945	2477914203	4033380444	3278587996	87849052	340695135	1026654305	1999904870	3581739114	616028266	...	2667063169	932734850	1785683843	2258843522	3846829958	1739267982	3206633359	3628883864	2513428378	3034212252	603510687	3995047843	2435680164	4278515623	848127915	3044751281	1135286194	1842145205	3454347189	3368912827	2238101436	2850656190	3981762496	2555238338	2088896453	2309124039	3523495880	152948679	717723595	458993620	3026372575	1534595042	2455228386	4223817698	1632503783	2037096428	2308325357	3011598321	4086265842	612622329
0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	3	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	1	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0
2	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	2	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
3	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
4	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	2	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
628	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
629	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	1	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
630	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
631	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
632	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0

633 rows × 693 columns

# FCFP2

	0	1	2	3	4	5	6	728953866	613541904	16	18	19	17	3764335124	20	614174231	3205495832	729008665	32	728943649	606136866	614173218	729164322	613541923	614173220	729164320	3766531108	594405931	728943668	614173237	728943675	614173243	3205495869	614173247	3766522948	613542983	613542989	3205495887	3205495891	614174292	...	613595558	614173606	728928169	594405804	728944045	594405807	3766523827	3766533048	594405818	614198715	594405820	613540798	613659071	3205496772	3764348868	3768571846	594910150	3766523334	729011661	594910160	728945618	1541843420	728497630	3208849886	3764344800	3764344801	728941536	613542373	613660134	3208849897	3208849903	614173680	3205493745	3764348914	3208849907	614173686	3764344823	3205496824	3205496825	614173690
0	15	0	1	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	3	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0
1	4	0	0	1	6	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	4	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	2	0	0	0	0	0	0	0	0	0	0	0	0	1	0
2	3	0	0	3	6	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	3	0	0	2	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	2	0	0	0	0	0	0	0	0	0	0	0	0	2	0
3	4	0	0	1	6	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	2	0	0	0	0	0	0	0	...	0	0	0	2	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	3	0	0	0	0	0	0	0	0	0	0	0	0	0	0
4	15	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	4	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
628	1	0	0	4	6	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0
629	1	0	1	2	5	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	1	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	1	0	1	0
630	2	0	0	0	7	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	2	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	2	0	0	0	0	0	0	0	1	0	0	0	0	0	0
631	4	0	0	1	6	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	2	0	0	0	0	0	0	0	...	0	0	0	2	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	3	0	0	0	0	0	0	0	0	0	0	0	0	0	0
632	2	0	2	0	6	0	0	0	0	1	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	...	0	0	0	2	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	2	0	0	0	0	0	0	1	0	0	0	0	0	0	0

633 rows × 207 columns

# ECFP4

	3793461249	2105860100	3925172229	1036713990	3325788165	2805301255	4008337421	1040949262	3350945806	467730450	2458968089	452370461	670646308	3163127853	667918382	1470644272	1454104624	3599515696	4248715315	913334324	1787035702	1567088695	3602047030	3890528317	1446633534	1821515839	804536389	1191084102	912908359	1930379335	3415654472	557449285	3635585100	2386108492	3501482061	2909626449	1692958801	2082340947	2438226004	469172305	...	3907878801	2271084433	1742798737	3628883864	3445374872	2513428378	2329247646	667123616	580206500	643719077	3332751270	3867221932	3343122350	2298806190	1918025650	3626221494	35790774	4253597625	838934460	3981762496	3230031810	2829279174	963149774	2810568659	1031544788	2299404245	2495102935	2296168412	1456226268	2850848738	1577254885	173416422	2037096428	2308325357	3813294065	2610397171	2476687347	379527158	3324518393	137543679
0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
2	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
3	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
4	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
628	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0
629	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
630	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
631	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
632	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0

633 rows × 3128 columns

# FCFP4

	0	1	2	3	4	5	6	2885967875	1907154955	1212801037	2374672399	16	17	18	19	20	3279839263	32	2640642081	3274137634	3366518817	2533883944	174874666	4023132220	2558574653	4222517311	1977434178	2123636802	3274137677	1976852558	1849409615	4197064793	4064444508	631562334	2605817951	530399336	3391340690	3764805781	1335812246	4209934490	...	1465622370	2864267110	2317311856	3164340084	1924489080	1924489086	2592841601	3205496706	3205496708	3205496709	635608965	3205496710	3205496712	3205496711	3205496716	546602898	3538681752	672587674	2760531868	1512521629	3205496734	3937968028	3695083428	1320861611	3027918765	3923320750	3096362947	3205496772	84008901	3660226502	3021553606	3140321250	276463590	1263411178	4224516078	3020292087	3205496824	3205496825	3660226556	3660226557
0	15	0	1	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0
1	4	0	0	1	6	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0
2	3	0	0	3	6	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	2	0	0
3	4	0	0	1	6	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
4	15	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
628	1	0	0	4	6	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	4	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
629	1	0	1	2	5	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0
630	2	0	0	0	7	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
631	4	0	0	1	6	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
632	2	0	2	0	6	0	0	0	0	0	0	1	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0

633 rows × 1450 columns

# ECFP6

	3325788165	1036713990	2805301255	2765094923	369950735	467730450	3197763603	2177204247	3471671321	3214213158	3163127853	2302443567	983433264	1787035702	2414182457	557449285	1191084102	1692958801	469172305	408551507	2082340947	1075183701	3840147540	4033380444	81461341	1026654305	2213904488	4286316648	3581739114	3048505454	327286897	3469148274	2827387003	2031583355	3794894977	3526197379	2691465350	392429703	1096220809	1360035982	...	3009412898	3040411436	2119794482	2273017667	3909418820	2328067911	3776905034	3577675600	538902353	2389770068	585105244	953089885	4105666400	3819896681	972160873	1807941485	2209152882	3769728884	978321287	3103424394	916848527	3907878801	2271084433	1742798737	2329247646	667123616	2376695715	33685414	3343122350	2298806190	960430021	906887112	310116301	963149774	3067576278	1425670109	2850848738	1577254885	2610397171	1808531449
0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0
1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
2	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
3	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
4	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
628	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
629	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
630	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
631	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
632	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0

633 rows × 6508 columns

# FCFP6

	0	1	2	3	4	5	6	3894042626	2290106376	1776730117	2081652746	1907154955	394584067	1212801037	4120633351	2374672399	16	17	18	19	20	1753890836	3575062556	1030635551	32	2640642081	3274137634	791404579	1212301348	2275049506	3366518817	3279839263	2533883944	174874666	396320811	2196316204	2669895726	1728970801	2069250099	3359506483	...	4064985010	3426910138	84549562	1972658111	2018164673	270032835	3205496772	3096362947	3660226502	3021553606	582639560	3514515398	668172231	84008901	797736906	4262903757	1227366354	123813844	49889236	1909891030	2145148886	4237549528	2970853337	670048219	3641950172	3141042141	3522355165	3140321250	2316976101	276463590	1263411178	4224516078	1609252846	3020292087	3205496824	3205496825	168894458	3660226556	3660226557	3529482238
0	15	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0
1	4	0	0	1	6	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0
2	3	0	0	3	6	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	2	0	0	0	0
3	4	0	0	1	6	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
4	15	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
628	1	0	0	4	6	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
629	1	0	1	2	5	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0
630	2	0	0	0	7	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
631	4	0	0	1	6	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
632	2	0	2	0	6	0	0	0	0	0	0	0	0	0	0	0	1	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0

633 rows × 4110 columns

いやー、まあ、こんなフィンガープリントのカラム名なんか見せられても、こんなん解釈困難やな。ほんまに。せやけど、こんなけぎょうさんの種類のフィンガープリントを簡単に計算できるように作ったんやで。偉いやろ。褒めて褒めて。

Mordred

記述子といえば Mordred が有名やんな。

%%time 
# mordredのインストール
!pip install mordred

Collecting mordred
  Downloading mordred-1.2.0.tar.gz (128 kB)
[K     |████████████████████████████████| 128 kB 5.1 MB/s 
[?25hRequirement already satisfied: six==1.* in /usr/local/lib/python3.7/dist-packages (from mordred) (1.15.0)
Requirement already satisfied: numpy==1.* in /usr/local/lib/python3.7/dist-packages (from mordred) (1.19.5)
Requirement already satisfied: networkx==2.* in /usr/local/lib/python3.7/dist-packages (from mordred) (2.6.3)
Building wheels for collected packages: mordred
  Building wheel for mordred (setup.py) ... [?25l[?25hdone
  Created wheel for mordred: filename=mordred-1.2.0-py3-none-any.whl size=176723 sha256=5b605782d809434f86e7473beeccbed4824da51a8e5b450715f81bc511d0b24c
  Stored in directory: /root/.cache/pip/wheels/02/c0/2e/e7e3d63b431777712ebc128bc4deb9ac5cb19afc7c1ea341ec
Successfully built mordred
Installing collected packages: mordred
Successfully installed mordred-1.2.0
CPU times: user 49.9 ms, sys: 13.4 ms, total: 63.3 ms
Wall time: 4.64 s

こんなふうに計算できんねんで。ちょっと時間かかっけどな。

%%time 
from rdkit import Chem
from mordred import Calculator, descriptors

calc = Calculator(descriptors)
mordred_descriptors_df = calc.pandas([Chem.MolFromSmiles(smile) for smile in df['Open Babel SMILES']])
display(mordred_descriptors_df)

100%|██████████| 633/633 [03:30<00:00,  3.01it/s]

	ABC	ABCGG	nAcid	nBase	SpAbs_A	SpMax_A	SpDiam_A	SpAD_A	SpMAD_A	LogEE_A	VE1_A	VE2_A	VE3_A	VR1_A	VR2_A	VR3_A	nAromAtom	nAromBond	nAtom	nHeavyAtom	nSpiro	nBridgehead	nHetero	nH	nB	nC	nN	nO	nS	nP	nF	nCl	nBr	nI	nX	ATS0dv	ATS1dv	ATS2dv	ATS3dv	ATS4dv	...	JGI7	JGI8	JGI9	JGI10	JGT10	Diameter	Radius	TopoShapeIndex	PetitjeanIndex	Vabc	VAdjMat	MWC01	MWC02	MWC03	MWC04	MWC05	MWC06	MWC07	MWC08	MWC09	MWC10	TMWC10	SRW02	SRW03	SRW04	SRW05	SRW06	SRW07	SRW08	SRW09	SRW10	TSRW10	MW	AMW	WPath	WPol	Zagreb1	Zagreb2	mZagreb1	mZagreb2
0	12.555834	11.602561	0	1	19.854843	2.352277	4.611329	19.854843	1.167932	3.715164	3.514302	0.206724	1.787469	129.529290	7.619370	5.394535	0	0	41	17	0	0	3	24	0	14	2	1	0	0	0	0	0	0	0	178.0	170.0	209.0	234.0	159.0	...	0.013021	0.015983	0.0	0.0	0.555706	9	5	0.800000	0.444444	261.218164	5.087463	17.0	4.394449	5.187386	6.037871	6.855409	7.711101	8.541495	9.398892	10.237278	11.095120	103.459001	3.555348	0.0	4.844187	0.000000	6.289716	2.70805	7.812378	5.379897	9.381348	56.970925	236.188863	5.760704	533	26	80.0	89.0	7.916667	3.833333
1	8.025464	7.826624	0	0	12.698196	2.326846	4.653693	12.698196	1.154381	3.295059	3.073114	0.279374	1.218002	37.827512	3.438865	3.728347	6	6	25	11	0	0	1	14	0	10	0	1	0	0	0	0	0	0	0	114.0	110.0	137.0	120.0	61.0	...	0.000000	0.000000	0.0	0.0	0.570105	6	3	1.000000	0.500000	159.140702	4.459432	11.0	3.970292	4.779123	5.638355	6.459904	7.323171	8.147578	9.011279	9.836439	10.700071	87.866213	3.135494	0.0	4.418841	0.000000	5.916202	0.00000	7.519150	0.000000	9.169831	41.159518	150.104465	6.004179	146	15	52.0	59.0	5.194444	2.555556
2	8.623181	8.454283	0	0	14.828323	2.340374	4.680749	14.828323	1.235694	3.376509	3.161639	0.263470	1.333412	45.534706	3.794559	4.000796	6	6	24	12	0	0	3	12	0	9	0	3	0	0	0	0	0	0	0	166.0	140.0	195.0	200.0	174.0	...	0.000000	0.000000	0.0	0.0	0.511503	6	4	0.500000	0.333333	159.425171	4.584963	12.0	4.043051	4.859812	5.713733	6.552508	7.407924	8.252446	9.106867	9.953277	10.806956	90.696575	3.218876	0.0	4.488636	0.000000	5.983936	0.00000	7.585281	0.000000	9.238442	42.515171	168.078644	7.003277	188	17	56.0	64.0	5.444444	2.888889
3	8.134854	7.770338	0	0	12.675204	2.302776	4.605551	12.675204	1.152291	3.294669	3.091013	0.281001	1.223809	37.212677	3.382971	3.711960	6	6	25	11	0	0	1	14	0	10	0	1	0	0	0	0	0	0	0	112.0	115.0	140.0	114.0	54.0	...	0.000000	0.000000	0.0	0.0	0.580361	6	3	1.000000	0.500000	159.140702	4.459432	11.0	3.970292	4.762174	5.609472	6.421622	7.271704	8.087948	8.938663	9.755741	10.606610	87.424225	3.135494	0.0	4.418841	0.000000	5.899897	0.00000	7.474772	0.000000	9.094144	41.023148	150.104465	6.004179	150	14	52.0	58.0	5.194444	2.472222
4	11.143219	9.394819	0	0	17.535313	2.233148	4.444177	17.535313	1.169021	3.590952	3.540074	0.236005	1.669613	66.874116	4.458274	4.608277	0	0	39	15	0	0	0	24	0	15	0	0	0	0	0	0	0	0	0	104.0	107.0	111.0	115.0	121.0	...	0.000000	0.000000	0.0	0.0	0.517928	6	5	0.200000	0.166667	247.730402	4.906891	15.0	4.262680	5.003946	5.831882	6.591674	7.418781	8.185629	9.011767	9.782844	10.608094	96.697298	3.433987	0.0	4.709530	0.000000	6.124683	0.00000	7.596392	0.000000	9.098403	45.962996	204.187801	5.235585	359	19	70.0	74.0	6.284722	3.333333
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
628	8.163363	8.029752	0	0	13.137460	2.364871	4.729742	13.137460	1.194315	3.303305	3.092493	0.281136	1.224288	35.727761	3.247978	3.671238	6	6	19	11	0	0	4	8	0	7	0	4	0	0	0	0	0	0	0	190.0	172.0	246.0	262.0	143.0	...	0.000000	0.000000	0.0	0.0	0.560975	5	3	0.666667	0.400000	133.623428	4.459432	11.0	4.007333	4.844187	5.707110	6.561031	7.424762	8.281724	9.145375	10.003107	10.866605	88.841233	3.135494	0.0	4.465908	0.000000	6.008813	0.00000	7.647309	0.000000	9.331052	41.588577	156.042259	8.212750	140	17	54.0	63.0	5.805556	2.444444
629	7.249407	6.952976	0	0	12.392927	2.262725	4.525450	12.392927	1.239293	3.197657	2.966570	0.296657	1.087406	33.218265	3.321827	3.503100	6	6	16	10	0	0	4	6	0	6	0	4	0	0	0	0	0	0	0	192.0	156.0	209.0	225.0	111.0	...	0.000000	0.000000	0.0	0.0	0.527680	6	3	1.000000	0.500000	116.327442	4.321928	10.0	3.850148	4.634729	5.451038	6.259581	7.077498	7.890957	8.708970	9.523617	10.341581	83.738119	3.044522	0.0	4.290459	0.000000	5.752573	0.00000	7.305860	0.000000	8.899867	39.293282	142.026609	8.876663	117	12	46.0	51.0	4.333333	2.361111
630	8.623181	8.048055	0	0	14.019917	2.385953	4.688728	14.019917	1.274538	3.352587	3.163201	0.287564	1.246895	40.693301	3.699391	3.801374	9	10	21	11	0	0	2	10	0	9	2	0	0	0	0	0	0	0	0	134.0	151.0	201.0	137.0	77.0	...	0.000000	0.000000	0.0	0.0	0.469370	5	3	0.666667	0.400000	129.155093	4.584963	12.0	4.077537	4.919981	5.789960	6.652863	7.523481	8.390949	9.261129	10.129866	10.999747	90.745514	3.218876	0.0	4.532599	2.397895	6.042633	4.59512	7.646354	6.580639	9.298809	55.312925	146.084398	6.956400	140	14	58.0	68.0	3.694444	2.388889
631	8.094413	7.861189	0	0	13.168697	2.318335	4.636669	13.168697	1.197154	3.295042	3.080399	0.280036	1.220369	37.279058	3.389005	3.713742	6	6	25	11	0	0	1	14	0	10	0	1	0	0	0	0	0	0	0	112.0	116.0	150.0	128.0	54.0	...	0.000000	0.000000	0.0	0.0	0.505208	5	3	0.666667	0.400000	159.140702	4.459432	11.0	3.970292	4.779123	5.624018	6.453625	7.299121	8.134761	8.978787	9.816785	10.659680	87.716192	3.135494	0.0	4.418841	0.000000	5.916202	0.00000	7.510431	0.000000	9.148784	41.129752	150.104465	6.004179	146	15	52.0	59.0	5.194444	2.500000
632	8.910910	8.632017	0	0	13.407823	2.358294	4.716589	13.407823	1.117319	3.383453	3.189168	0.265764	1.342082	42.532621	3.544385	3.932593	6	6	22	12	0	0	4	10	0	8	2	2	0	0	0	0	0	0	0	194.0	168.0	239.0	216.0	145.0	...	0.000000	0.000000	0.0	0.0	0.605111	6	3	1.000000	0.500000	152.696021	4.584963	12.0	4.077537	4.890349	5.771441	6.597146	7.482682	8.311153	9.197559	10.026634	10.913214	91.267715	3.218876	0.0	4.532599	0.000000	6.042633	0.00000	7.657755	0.000000	9.324115	42.775979	166.074228	7.548829	184	17	58.0	66.0	6.055556	2.666667

633 rows × 1826 columns

CPU times: user 30.6 s, sys: 239 ms, total: 30.8 s
Wall time: 45.7 s

Mordredは特徴量の数がめっちゃぎょうさんあんのが売りの一つや。その意味とかは公式ホームページの説明を読まはってや。

トレーニングデータ、テストデータへの分割

ほんでこれから、トレーニングデータ、テストデータに分割すっで。ここまで、いろんな記述子とかフィンガープリントを計算したやろ。そいつらが全部同じ分割になるようにすんねん。

import random

test_size = 0.2
ids = [id for id in range(len(df))]
random.shuffle(ids)
split_line = int(len(df) * (1 - test_size))
train_ids = ids[:split_line]
test_ids = ids[split_line:]

この train_ids がトレーニングデータのID、 test_ids がテストデータのIDっちゅうこっちゃな。

学習

ほな、まとめて Random Forest で学習してみんで。なんで Random Forest やねんっちゅうと、ハイパーパラメーターチューニングめんどいやろ？ ここでは単に記述子同士の比較がしたいだけやねんから、チューニングせえへんくてもそこそこの性能が出る言われてて、さらに特徴量の相対的な重要性が計算できる Random Forest が一番ええ思たねん。知らんけど。

import timeit
import pandas as pd
from sklearn.ensemble import RandomForestRegressor

from rdkit_installer.preprocess import TableCleaner
from rdkit_installer.depict import rf_feature_importances

def make_comparison(X_names, Xs, Y):
    result = {"train score":[], "test score":[], "train shape":[], "test shape":[], "time":[]}

    for name, X in zip(X_names, Xs):

        # エラーを起こす行とか列とか除外するで
        cleaner = TableCleaner()
        selected_col = cleaner.clean_columns(X, Y)
        selected_row = cleaner.clean_rows(X)
        train_set = list(set(train_ids) & set(selected_row))
        test_set = list(set(test_ids) & set(selected_row))
        X_train, X_test = X.iloc[train_set, selected_col], X.iloc[test_set, selected_col]
        y_train, y_test = Y.iloc[train_set, :], Y.iloc[test_set, :]

        # 学習しながら計算時間も計測すんねん
        regr = RandomForestRegressor()
        seconds = timeit.timeit(lambda: regr.fit(X_train, y_train.values.ravel()), number=1)

        # 計算結果（性能）を記録するんや
        result["time"].append(seconds) # 学習に要した時間（秒）
        result["train shape"].append(X_train.shape)
        result["test shape"].append(X_test.shape)
        result["train score"].append(regr.score(X_train, y_train.values.ravel()))
        result["test score"].append(regr.score(X_test, y_test.values.ravel()))

        # 貢献度の高い特徴量トップ１０を図示すんで
        print(name, "Test score=", regr.score(X_test, y_test.values.ravel()))
        rf_feature_importances(regr, X)

    # pandas 形式にしといたほうが後で見やすいさかい、そうしとくな
    data = pd.DataFrame(result)
    data.index = X_names
    return data

"HOMO-LUMO gap" を目的変数とした回帰モデルをまとめて構築しはんねんたら、こんな感じや。

%%time 
result1 = make_comparison(
    ["RDKit 208 descriptors", "RDKit 400 descriptors", "Mordred 1826 descriptors"],
    [rdkit_208descriptors_df, rdkit_descriptors_df, mordred_descriptors_df], 
    df[["HOMO-LUMO gap"]]
    )

RDKit 208 descriptors Test score= 0.7393261212344209

RDKit 400 descriptors Test score= 0.7508808273760392

Mordred 1826 descriptors Test score= 0.7116458690779182

CPU times: user 47.3 s, sys: 2.2 s, total: 49.5 s
Wall time: 3min 37s

データ分割はランダムにしてんねんけど、それによって順位が変わることがあんねん。せやけどまあ、上の例だけで言うたら RDKit の４００個の記述子が一番性能が良かったみたいやな。そうは言うけど、さっきも言うた通り、特徴量を解釈しよう思わはるねんたら、RDKitに新しく追加された記述子は、難ありなんや。まあ知らんけど。

全体的な性能について、表にまとめたらこんな感じや。

result1

	train score	test score	train shape	test shape	time
RDKit 208 descriptors	0.964067	0.739326	(506, 208)	(127, 208)	2.123547
RDKit 400 descriptors	0.963012	0.750881	(496, 320)	(123, 320)	3.914356
Mordred 1826 descriptors	0.957463	0.711646	(506, 1278)	(127, 1278)	19.107194

ついでやし、各種フィンガープリントについても計算してみてん。貢献度の高い特徴量トップ１０も出力されるけど、こんなん解釈せえ言わはっても困難やで。ほんまに。

%%time
result2 = make_comparison(
    fingerprinter.names,
    fp_dfs, 
    df[["HOMO-LUMO gap"]]
    )

MACCSkeys Test score= 0.7428648213992282

Avalon Test score= 0.6600486101886427

Morgan2(1024bits) Test score= 0.564839806463383

Morgan2F(1024bits) Test score= 0.6145275706762224

Morgan4(2048bits) Test score= 0.4589796057137068

Morgan4F(2048bits) Test score= 0.5606490833767938

ECFP2 Test score= 0.6797667810884447

FCFP2 Test score= 0.6089529501171054

ECFP4 Test score= 0.6816544563865634

FCFP4 Test score= 0.6067634726762778

ECFP6 Test score= 0.6805538722932313

FCFP6 Test score= 0.6192763399924386

CPU times: user 41.4 s, sys: 205 ms, total: 41.6 s
Wall time: 42.8 s

各フィンガープリントから得られた性能を表にまとめると、こんな感じや。うひょー。

result2

	train score	test score	train shape	test shape	time
MACCSkeys	0.955100	0.742865	(506, 167)	(127, 167)	0.569669
Avalon	0.954562	0.660049	(506, 512)	(127, 512)	1.899150
Morgan2(1024bits)	0.949939	0.564840	(506, 1024)	(127, 1024)	2.130418
Morgan2F(1024bits)	0.952999	0.614528	(506, 1024)	(127, 1024)	1.596574
Morgan4(2048bits)	0.944960	0.458980	(506, 2048)	(127, 2048)	4.163680
Morgan4F(2048bits)	0.947081	0.560649	(506, 2048)	(127, 2048)	3.517748
ECFP2	0.961705	0.679767	(506, 693)	(127, 693)	1.167866
FCFP2	0.948420	0.608953	(506, 207)	(127, 207)	0.576670
ECFP4	0.957817	0.681654	(506, 3128)	(127, 3128)	3.668387
FCFP4	0.956894	0.606763	(506, 1450)	(127, 1450)	1.988229
ECFP6	0.963806	0.680554	(506, 6508)	(127, 6508)	6.774919
FCFP6	0.958517	0.619276	(506, 4110)	(127, 4110)	4.656614

いちおう、記述子の結果も再掲しとくで。

result1

	train score	test score	train shape	test shape	time
RDKit 208 descriptors	0.964067	0.739326	(506, 208)	(127, 208)	2.123547
RDKit 400 descriptors	0.963012	0.750881	(496, 320)	(123, 320)	3.914356
Mordred 1826 descriptors	0.957463	0.711646	(506, 1278)	(127, 1278)	19.107194

結果の総括

こんなん数字だけ見せられても比較困難やわ、って声が聞こえてきたさかい、プロットしてみんで。

import matplotlib.pyplot as plt

data = pd.concat([result1, result2])
for colname1 in ["time", "test score"]:
    for colname2 in ["test score", "train score"]:
        if colname1 != colname2:
            X = list(data[colname1])
            Y = list(data[colname2])
            plt.figure(figsize=(8,8))
            plt.scatter(X, Y)
            for x, y, name in zip(X, Y, data.index):
                plt.text(x, y, name, alpha=0.8, size=8)
            plt.xlabel(colname1)
            plt.ylabel(colname2)
            plt.grid()
            plt.show()

まとめたら、Mordred は計算時間がかかる割にはベストな性能っちゅうわけでもあらへんくて、RDKit の ４００個の記述子が、計算時間の面でも性能の面でも一番ええように見えるな。あとは、Morganよりも、FCFPよりも、ECFPのほうが頑張らはったみたいや。MACCS keys とか Avalon とか、意外と健闘してはんねんな。興味ある人いはったら詳しく検討しはったらええで。

今回検討したんは１例だけやけど、データセットの分割とか、目的変数によって、違った結果が得られることは充分にありえると思うで。知らんけど。

あとはまあ恥ずかしい話やけど、フィンガープリントの定義とか計算方法とか間違っている部分があるかも知れへん。ほんま堪忍やで。根幹部分のコードは https://github.com/maskot1977/rdkit_installer に記述しといたから、何や気付かはった点があったら指摘してくれはるとほんま助かります。

Take Home Message

RDKit の新機能、まだまだあるでー、きっと。

RDKit だけに。