Feature

A collection of tools to extract features from SMILES, proteins, etc.

Setup

Features from SMILES

RDKit descriptors

---

source

get_rdkit

 get_rdkit (df:pandas.core.frame.DataFrame, col:str='SMILES',
            normalize:bool=True)

Extract chemical features from smiles via rdkit.Chem.Descriptors; if normalize, apply StandardScaler

	Type	Default	Details
df	DataFrame		a dataframe that contains smiles
col	str	SMILES	colname of smile
normalize	bool	True	normalize features using StandardScaler()

aa = Data.get_aa_info()
aa.head()

	Name	SMILES	MW	pKa1	pKb2	pKx3	pl4	H	VSC	P1	P2	SASA	NCISC	phospho
aa
A	Alanine	C[C@@H](C(=O)O)N	89.10	2.34	9.69	NaN	6.00	0.62	27.5	8.1	0.046	1.181	0.007187	0
C	Cysteine	C([C@@H](C(=O)O)N)S	121.16	1.96	10.28	8.18	5.07	0.29	44.6	5.5	0.128	1.461	-0.036610	0
D	Aspartic acid	C([C@@H](C(=O)O)N)C(=O)O	133.11	1.88	9.60	3.65	2.77	-0.90	40.0	13.0	0.105	1.587	-0.023820	0
E	Glutamic acid	C(CC(=O)O)[C@@H](C(=O)O)N	147.13	2.19	9.67	4.25	3.22	-0.74	62.0	12.3	0.151	1.862	0.006802	0
F	Phenylalanine	c1ccc(cc1)C[C@@H](C(=O)O)N	165.19	1.83	9.13	NaN	5.48	1.19	115.5	5.2	0.290	2.228	0.037552	0

aa_rdkit = get_rdkit(aa, 'SMILES')
aa_rdkit.head()

	MaxAbsEStateIndex	MaxEStateIndex	MinAbsEStateIndex	MinEStateIndex	qed	SPS	MolWt	HeavyAtomMolWt	ExactMolWt	NumValenceElectrons	...	fr_sulfide	fr_sulfonamd	fr_sulfone	fr_term_acetylene	fr_tetrazole	fr_thiazole	fr_thiocyan	fr_thiophene	fr_unbrch_alkane	fr_urea
aa
A	-1.653421	-1.653421	1.218945	0.407753	-0.383393	-0.070345	-1.523488	-1.498443	-1.523163	-1.551232	...	-0.204124	0.0	0.0	0.0	0.0	0.0	0.0	0.0	-0.408248	0.0
C	-1.058215	-1.058215	-0.588000	0.372307	-0.641865	-0.138884	-0.727067	-0.666442	-0.728722	-1.137202	...	-0.204124	0.0	0.0	0.0	0.0	0.0	0.0	0.0	-0.408248	0.0
D	-0.764466	-0.764466	0.554854	0.126078	-0.376981	-0.390192	-0.430473	-0.356599	-0.430106	-0.447152	...	-0.204124	0.0	0.0	0.0	0.0	0.0	0.0	0.0	-0.408248	0.0
E	-0.283221	-0.283221	-1.143984	0.235448	-0.051582	-0.406185	-0.082096	-0.044965	-0.081845	-0.033122	...	-0.204124	0.0	0.0	0.0	0.0	0.0	0.0	0.0	-0.408248	0.0
F	0.972596	0.972596	0.063427	0.410788	1.908107	-0.430173	0.366494	0.371360	0.366059	0.380907	...	-0.204124	0.0	0.0	0.0	0.0	0.0	0.0	0.0	-0.408248	0.0

5 rows × 210 columns

Morgan fingerprint

---

source

get_morgan

 get_morgan (df:pandas.core.frame.DataFrame, col:str='SMILES', radius=3)

Get 2048 morgan fingerprint (binary feature) from smiles in a dataframe

	Type	Default	Details
df	DataFrame		a dataframe that contains smiles
col	str	SMILES	colname of smile
radius	int	3

aa_morgan = get_morgan(aa, 'SMILES')
aa_morgan.head()

	morgan_0	morgan_1	morgan_2	morgan_3	morgan_4	morgan_5	morgan_6	morgan_7	morgan_8	morgan_9	morgan_10	morgan_11	morgan_12	morgan_13	morgan_14	morgan_15	morgan_16	morgan_17	morgan_18	morgan_19	morgan_20	morgan_21	morgan_22	morgan_23	morgan_24	morgan_25	morgan_26	morgan_27	morgan_28	morgan_29	morgan_30	morgan_31	morgan_32	morgan_33	morgan_34	morgan_35	morgan_36	morgan_37	morgan_38	morgan_39	morgan_40	morgan_41	morgan_42	morgan_43	morgan_44	morgan_45	morgan_46	morgan_47	morgan_48	morgan_49	morgan_50	morgan_51	morgan_52	morgan_53	morgan_54	morgan_55	morgan_56	morgan_57	morgan_58	morgan_59	morgan_60	morgan_61	morgan_62	morgan_63	morgan_64	morgan_65	morgan_66	morgan_67	morgan_68	morgan_69	morgan_70	morgan_71	morgan_72	morgan_73	morgan_74	morgan_75	morgan_76	morgan_77	morgan_78	morgan_79	morgan_80	morgan_81	morgan_82	morgan_83	morgan_84	morgan_85	morgan_86	morgan_87	morgan_88	morgan_89	morgan_90	morgan_91	morgan_92	morgan_93	morgan_94	morgan_95	morgan_96	morgan_97	morgan_98	morgan_99	morgan_100	morgan_101	morgan_102	morgan_103	morgan_104	morgan_105	morgan_106	morgan_107	morgan_108	morgan_109	morgan_110	morgan_111	morgan_112	morgan_113	morgan_114	morgan_115	morgan_116	morgan_117	morgan_118	morgan_119	morgan_120	morgan_121	morgan_122	morgan_123	morgan_124	morgan_125	morgan_126	morgan_127	morgan_128	morgan_129	morgan_130	morgan_131	morgan_132	morgan_133	morgan_134	morgan_135	morgan_136	morgan_137	morgan_138	morgan_139	morgan_140	morgan_141	morgan_142	morgan_143	morgan_144	morgan_145	morgan_146	morgan_147	morgan_148	morgan_149	morgan_150	morgan_151	morgan_152	morgan_153	morgan_154	morgan_155	morgan_156	morgan_157	morgan_158	morgan_159	morgan_160	morgan_161	morgan_162	morgan_163	morgan_164	morgan_165	morgan_166	morgan_167	morgan_168	morgan_169	morgan_170	morgan_171	morgan_172	morgan_173	morgan_174	morgan_175	morgan_176	morgan_177	morgan_178	morgan_179	morgan_180	morgan_181	morgan_182	morgan_183	morgan_184	morgan_185	morgan_186	morgan_187	morgan_188	morgan_189	morgan_190	morgan_191	morgan_192	morgan_193	morgan_194	morgan_195	morgan_196	morgan_197	morgan_198	morgan_199	morgan_200	morgan_201	morgan_202	morgan_203	morgan_204	morgan_205	morgan_206	morgan_207	morgan_208	morgan_209	morgan_210	morgan_211	morgan_212	morgan_213	morgan_214	morgan_215	morgan_216	morgan_217	morgan_218	morgan_219	morgan_220	morgan_221	morgan_222	morgan_223	morgan_224	morgan_225	morgan_226	morgan_227	morgan_228	morgan_229	morgan_230	morgan_231	morgan_232	morgan_233	morgan_234	morgan_235	morgan_236	morgan_237	morgan_238	morgan_239	morgan_240	morgan_241	morgan_242	morgan_243	morgan_244	morgan_245	morgan_246	morgan_247	morgan_248	morgan_249	morgan_250	morgan_251	morgan_252	morgan_253	morgan_254	morgan_255	morgan_256	morgan_257	morgan_258	morgan_259	morgan_260	morgan_261	morgan_262	morgan_263	morgan_264	morgan_265	morgan_266	morgan_267	morgan_268	morgan_269	morgan_270	morgan_271	morgan_272	morgan_273	morgan_274	morgan_275	morgan_276	morgan_277	morgan_278	morgan_279	morgan_280	morgan_281	morgan_282	morgan_283	morgan_284	morgan_285	morgan_286	morgan_287	morgan_288	morgan_289	morgan_290	morgan_291	morgan_292	morgan_293	morgan_294	morgan_295	morgan_296	morgan_297	morgan_298	morgan_299	morgan_300	morgan_301	morgan_302	morgan_303	morgan_304	morgan_305	morgan_306	morgan_307	morgan_308	morgan_309	morgan_310	morgan_311	morgan_312	morgan_313	morgan_314	morgan_315	morgan_316	morgan_317	morgan_318	morgan_319	morgan_320	morgan_321	morgan_322	morgan_323	morgan_324	morgan_325	morgan_326	morgan_327	morgan_328	morgan_329	morgan_330	morgan_331	morgan_332	morgan_333	morgan_334	morgan_335	morgan_336	morgan_337	morgan_338	morgan_339	morgan_340	morgan_341	morgan_342	morgan_343	morgan_344	morgan_345	morgan_346	morgan_347	morgan_348	morgan_349	morgan_350	morgan_351	morgan_352	morgan_353	morgan_354	morgan_355	morgan_356	morgan_357	morgan_358	morgan_359	morgan_360	morgan_361	morgan_362	morgan_363	morgan_364	morgan_365	morgan_366	morgan_367	morgan_368	morgan_369	morgan_370	morgan_371	morgan_372	morgan_373	morgan_374	morgan_375	morgan_376	morgan_377	morgan_378	morgan_379	morgan_380	morgan_381	morgan_382	morgan_383	morgan_384	morgan_385	morgan_386	morgan_387	morgan_388	morgan_389	morgan_390	morgan_391	morgan_392	morgan_393	morgan_394	morgan_395	morgan_396	morgan_397	morgan_398	morgan_399	morgan_400	morgan_401	morgan_402	morgan_403	morgan_404	morgan_405	morgan_406	morgan_407	morgan_408	morgan_409	morgan_410	morgan_411	morgan_412	morgan_413	morgan_414	morgan_415	morgan_416	morgan_417	morgan_418	morgan_419	morgan_420	morgan_421	morgan_422	morgan_423	morgan_424	morgan_425	morgan_426	morgan_427	morgan_428	morgan_429	morgan_430	morgan_431	morgan_432	morgan_433	morgan_434	morgan_435	morgan_436	morgan_437	morgan_438	morgan_439	morgan_440	morgan_441	morgan_442	morgan_443	morgan_444	morgan_445	morgan_446	morgan_447	morgan_448	morgan_449	morgan_450	morgan_451	morgan_452	morgan_453	morgan_454	morgan_455	morgan_456	morgan_457	morgan_458	morgan_459	morgan_460	morgan_461	morgan_462	morgan_463	morgan_464	morgan_465	morgan_466	morgan_467	morgan_468	morgan_469	morgan_470	morgan_471	morgan_472	morgan_473	morgan_474	morgan_475	morgan_476	morgan_477	morgan_478	morgan_479	morgan_480	morgan_481	morgan_482	morgan_483	morgan_484	morgan_485	morgan_486	morgan_487	morgan_488	morgan_489	morgan_490	morgan_491	morgan_492	morgan_493	morgan_494	morgan_495	morgan_496	morgan_497	morgan_498	...	morgan_1549	morgan_1550	morgan_1551	morgan_1552	morgan_1553	morgan_1554	morgan_1555	morgan_1556	morgan_1557	morgan_1558	morgan_1559	morgan_1560	morgan_1561	morgan_1562	morgan_1563	morgan_1564	morgan_1565	morgan_1566	morgan_1567	morgan_1568	morgan_1569	morgan_1570	morgan_1571	morgan_1572	morgan_1573	morgan_1574	morgan_1575	morgan_1576	morgan_1577	morgan_1578	morgan_1579	morgan_1580	morgan_1581	morgan_1582	morgan_1583	morgan_1584	morgan_1585	morgan_1586	morgan_1587	morgan_1588	morgan_1589	morgan_1590	morgan_1591	morgan_1592	morgan_1593	morgan_1594	morgan_1595	morgan_1596	morgan_1597	morgan_1598	morgan_1599	morgan_1600	morgan_1601	morgan_1602	morgan_1603	morgan_1604	morgan_1605	morgan_1606	morgan_1607	morgan_1608	morgan_1609	morgan_1610	morgan_1611	morgan_1612	morgan_1613	morgan_1614	morgan_1615	morgan_1616	morgan_1617	morgan_1618	morgan_1619	morgan_1620	morgan_1621	morgan_1622	morgan_1623	morgan_1624	morgan_1625	morgan_1626	morgan_1627	morgan_1628	morgan_1629	morgan_1630	morgan_1631	morgan_1632	morgan_1633	morgan_1634	morgan_1635	morgan_1636	morgan_1637	morgan_1638	morgan_1639	morgan_1640	morgan_1641	morgan_1642	morgan_1643	morgan_1644	morgan_1645	morgan_1646	morgan_1647	morgan_1648	morgan_1649	morgan_1650	morgan_1651	morgan_1652	morgan_1653	morgan_1654	morgan_1655	morgan_1656	morgan_1657	morgan_1658	morgan_1659	morgan_1660	morgan_1661	morgan_1662	morgan_1663	morgan_1664	morgan_1665	morgan_1666	morgan_1667	morgan_1668	morgan_1669	morgan_1670	morgan_1671	morgan_1672	morgan_1673	morgan_1674	morgan_1675	morgan_1676	morgan_1677	morgan_1678	morgan_1679	morgan_1680	morgan_1681	morgan_1682	morgan_1683	morgan_1684	morgan_1685	morgan_1686	morgan_1687	morgan_1688	morgan_1689	morgan_1690	morgan_1691	morgan_1692	morgan_1693	morgan_1694	morgan_1695	morgan_1696	morgan_1697	morgan_1698	morgan_1699	morgan_1700	morgan_1701	morgan_1702	morgan_1703	morgan_1704	morgan_1705	morgan_1706	morgan_1707	morgan_1708	morgan_1709	morgan_1710	morgan_1711	morgan_1712	morgan_1713	morgan_1714	morgan_1715	morgan_1716	morgan_1717	morgan_1718	morgan_1719	morgan_1720	morgan_1721	morgan_1722	morgan_1723	morgan_1724	morgan_1725	morgan_1726	morgan_1727	morgan_1728	morgan_1729	morgan_1730	morgan_1731	morgan_1732	morgan_1733	morgan_1734	morgan_1735	morgan_1736	morgan_1737	morgan_1738	morgan_1739	morgan_1740	morgan_1741	morgan_1742	morgan_1743	morgan_1744	morgan_1745	morgan_1746	morgan_1747	morgan_1748	morgan_1749	morgan_1750	morgan_1751	morgan_1752	morgan_1753	morgan_1754	morgan_1755	morgan_1756	morgan_1757	morgan_1758	morgan_1759	morgan_1760	morgan_1761	morgan_1762	morgan_1763	morgan_1764	morgan_1765	morgan_1766	morgan_1767	morgan_1768	morgan_1769	morgan_1770	morgan_1771	morgan_1772	morgan_1773	morgan_1774	morgan_1775	morgan_1776	morgan_1777	morgan_1778	morgan_1779	morgan_1780	morgan_1781	morgan_1782	morgan_1783	morgan_1784	morgan_1785	morgan_1786	morgan_1787	morgan_1788	morgan_1789	morgan_1790	morgan_1791	morgan_1792	morgan_1793	morgan_1794	morgan_1795	morgan_1796	morgan_1797	morgan_1798	morgan_1799	morgan_1800	morgan_1801	morgan_1802	morgan_1803	morgan_1804	morgan_1805	morgan_1806	morgan_1807	morgan_1808	morgan_1809	morgan_1810	morgan_1811	morgan_1812	morgan_1813	morgan_1814	morgan_1815	morgan_1816	morgan_1817	morgan_1818	morgan_1819	morgan_1820	morgan_1821	morgan_1822	morgan_1823	morgan_1824	morgan_1825	morgan_1826	morgan_1827	morgan_1828	morgan_1829	morgan_1830	morgan_1831	morgan_1832	morgan_1833	morgan_1834	morgan_1835	morgan_1836	morgan_1837	morgan_1838	morgan_1839	morgan_1840	morgan_1841	morgan_1842	morgan_1843	morgan_1844	morgan_1845	morgan_1846	morgan_1847	morgan_1848	morgan_1849	morgan_1850	morgan_1851	morgan_1852	morgan_1853	morgan_1854	morgan_1855	morgan_1856	morgan_1857	morgan_1858	morgan_1859	morgan_1860	morgan_1861	morgan_1862	morgan_1863	morgan_1864	morgan_1865	morgan_1866	morgan_1867	morgan_1868	morgan_1869	morgan_1870	morgan_1871	morgan_1872	morgan_1873	morgan_1874	morgan_1875	morgan_1876	morgan_1877	morgan_1878	morgan_1879	morgan_1880	morgan_1881	morgan_1882	morgan_1883	morgan_1884	morgan_1885	morgan_1886	morgan_1887	morgan_1888	morgan_1889	morgan_1890	morgan_1891	morgan_1892	morgan_1893	morgan_1894	morgan_1895	morgan_1896	morgan_1897	morgan_1898	morgan_1899	morgan_1900	morgan_1901	morgan_1902	morgan_1903	morgan_1904	morgan_1905	morgan_1906	morgan_1907	morgan_1908	morgan_1909	morgan_1910	morgan_1911	morgan_1912	morgan_1913	morgan_1914	morgan_1915	morgan_1916	morgan_1917	morgan_1918	morgan_1919	morgan_1920	morgan_1921	morgan_1922	morgan_1923	morgan_1924	morgan_1925	morgan_1926	morgan_1927	morgan_1928	morgan_1929	morgan_1930	morgan_1931	morgan_1932	morgan_1933	morgan_1934	morgan_1935	morgan_1936	morgan_1937	morgan_1938	morgan_1939	morgan_1940	morgan_1941	morgan_1942	morgan_1943	morgan_1944	morgan_1945	morgan_1946	morgan_1947	morgan_1948	morgan_1949	morgan_1950	morgan_1951	morgan_1952	morgan_1953	morgan_1954	morgan_1955	morgan_1956	morgan_1957	morgan_1958	morgan_1959	morgan_1960	morgan_1961	morgan_1962	morgan_1963	morgan_1964	morgan_1965	morgan_1966	morgan_1967	morgan_1968	morgan_1969	morgan_1970	morgan_1971	morgan_1972	morgan_1973	morgan_1974	morgan_1975	morgan_1976	morgan_1977	morgan_1978	morgan_1979	morgan_1980	morgan_1981	morgan_1982	morgan_1983	morgan_1984	morgan_1985	morgan_1986	morgan_1987	morgan_1988	morgan_1989	morgan_1990	morgan_1991	morgan_1992	morgan_1993	morgan_1994	morgan_1995	morgan_1996	morgan_1997	morgan_1998	morgan_1999	morgan_2000	morgan_2001	morgan_2002	morgan_2003	morgan_2004	morgan_2005	morgan_2006	morgan_2007	morgan_2008	morgan_2009	morgan_2010	morgan_2011	morgan_2012	morgan_2013	morgan_2014	morgan_2015	morgan_2016	morgan_2017	morgan_2018	morgan_2019	morgan_2020	morgan_2021	morgan_2022	morgan_2023	morgan_2024	morgan_2025	morgan_2026	morgan_2027	morgan_2028	morgan_2029	morgan_2030	morgan_2031	morgan_2032	morgan_2033	morgan_2034	morgan_2035	morgan_2036	morgan_2037	morgan_2038	morgan_2039	morgan_2040	morgan_2041	morgan_2042	morgan_2043	morgan_2044	morgan_2045	morgan_2046	morgan_2047
aa
A	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	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	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	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	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	0	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	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	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	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	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	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
C	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	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	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	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	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	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	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	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	0	0	0	0	0	0	0	0	0	0	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	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	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
D	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	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	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	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	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	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	0	0	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	0	0	0	0	0	0	0	0	0	0	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	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
E	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	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	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	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	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	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	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	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	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	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	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	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
F	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	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	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	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	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	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	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	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	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	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	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0

5 rows × 2048 columns

aa_morgan = get_morgan(aa, 'SMILES')
aa_morgan.head()

	morgan_0	morgan_1	morgan_2	morgan_3	morgan_4	morgan_5	morgan_6	morgan_7	morgan_8	morgan_9	...	morgan_2038	morgan_2039	morgan_2040	morgan_2041	morgan_2042	morgan_2043	morgan_2044	morgan_2045	morgan_2046	morgan_2047
aa
A	0	1	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0
C	0	1	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0
D	0	1	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0
E	0	1	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0
F	0	1	0	0	0	0	0	0	0	0	...	0	0	0	0	0	0	0	0	0	0

5 rows × 2048 columns

Features from protein sequence

ESM2

---

source

get_esm

 get_esm (df:pandas.core.frame.DataFrame, col:str='sequence',
          model_name:str='esm2_t33_650M_UR50D')

Extract esmfold2 embeddings from protein sequence in a dataframe

	Type	Default	Details
df	DataFrame		a dataframe that contains amino acid sequence
col	str	sequence	colname of amino acid sequence
model_name	str	esm2_t33_650M_UR50D	Name of the ESM model to use for the embeddings.

ESM2 model is trained on UniRef sequence. The default model in the function is esm2_t33_650M_UR50D, which is trained on UniRef50.

Uncheck below to use:

# # Examples
# df = Data.get_kinase_info().set_index('kinase')
# sample = df[:5]
# esmfeature = get_esm(sample,'sequence')
# esmfeature.head()

ProtT5

---

source

get_t5

 get_t5 (df:pandas.core.frame.DataFrame, col:str='sequence')

Extract ProtT5-XL-uniref50 embeddings from protein sequence in a dataframe

XL-uniref50 model is a t5-3b model trained on Uniref50 Dataset.

Uncheck below to use:

# t5feature = get_t5(sample,'sequence')
# t5feature.head()

---

source

get_t5_bfd

 get_t5_bfd (df:pandas.core.frame.DataFrame, col:str='sequence')

Extract ProtT5-XL-BFD embeddings from protein sequence in a dataframe

XL-BFD model is a t5-3b model trained on Big Fantastic Database(BFD).

Uncheck below to use:

# t5bfd = get_t5_bfd(sample,'sequence')
# t5bfd.head()

Dimensionality reduction

---

source

reduce_feature

 reduce_feature (df:pandas.core.frame.DataFrame, method:str='pca',
                 complexity:int=20, n:int=2, load:str=None, save:str=None,
                 seed:int=123, **kwargs)

Reduce the dimensionality given a dataframe of values

	Type	Default	Details
df	DataFrame
method	str	pca	dimensionality reduction method, accept both capital and lower case
complexity	int	20	None for PCA; perfplexity for TSNE, recommend: 30; n_neigbors for UMAP, recommend: 15
n	int	2	n_components
load	str	None	load a previous model, e.g. model.pkl
save	str	None	pkl file to be saved, e.g. pca_model.pkl
seed	int	123	seed for random_state
kwargs	VAR_KEYWORD

A very common way to reduce feature number is to use dimensionality reduction method. reduce_feature is a dimensionality reduction function that can apply three dimensionality reduction methods: PCA, UMAP, TSNE. The later two is non-linear transformation, and PCA is linear transformation. Therefore, for plotting purpose, it is good to use UMAP/TSNE, by setting n (n_components) to 2 for 2d plot; for featuring purpose, it is good to use PCA, and set n to values to a rational values, like 64, 128 etc.

# Load data
df = Data.get_aa_rdkit()

# Use PCA to reduce dimension; reduce the number of features to 20
reduce_feature(df,'pca',n=20).head()

	PCA1	PCA2	PCA3	PCA4	PCA5	PCA6	PCA7	PCA8	PCA9	PCA10	PCA11	PCA12	PCA13	PCA14	PCA15	PCA16	PCA17	PCA18	PCA19	PCA20
aa
A	-463.014948	-79.180061	-8.957621	-13.455810	1.334975	0.996915	-6.228205	-2.573987	0.637178	1.904173	-1.165818	5.803809	-3.519867	1.620306	-3.686674	-1.070729	-1.044587	-2.245493	-0.023173	2.143434
C	-446.251885	-52.851228	1.200874	0.469406	-16.721236	12.310611	5.623647	17.543569	6.290376	-4.818617	0.871101	-1.274344	3.983329	-6.019231	-5.866159	-0.339787	-3.342606	0.934348	-0.335121	-0.045040
D	-407.721016	9.532878	10.375789	-21.871983	-3.757091	-2.804468	3.684495	-8.257556	0.885011	4.454468	4.085862	3.059634	3.463971	2.626308	1.260286	-3.054156	-1.823627	4.300302	-2.407938	1.076088
E	-355.786380	21.077202	11.870110	-10.861780	4.869825	-3.906521	-2.281413	-2.893303	8.997722	3.828554	2.004998	-1.002484	9.471326	-1.945113	3.684237	2.119809	0.362597	-1.166347	-3.604024	-0.752169
F	69.598210	74.375112	-68.407808	2.572185	6.659703	16.787547	10.585299	-1.588954	-10.532959	-2.643261	-7.012191	4.522682	4.779671	0.908865	-0.218389	1.152286	0.560021	1.736196	0.152801	-0.567469

---

source

remove_hi_corr

 remove_hi_corr (df:pandas.core.frame.DataFrame, thr:float=0.98)

Remove highly correlated features in a dataframe given a pearson threshold

	Type	Default	Details
df	DataFrame
thr	float	0.98	threshold

Another way to reduce features besides PCA/UMAP/TSNE is to remove features that are highly correlated. remove_hi_corr is a function to remove highly correlated features based on threshold of Pearson correlation between features.

# Load data
df = Data.get_aa_rdkit()
df.shape

(25, 106)

remove_hi_corr(df,thr=0.9).shape

(25, 78)

---

source

preprocess

 preprocess (df:pandas.core.frame.DataFrame, thr:float=0.98)

Remove features with no variance, and highly correlated features based on threshold

This function is similar to remove_hi_corr, but can additionaly remove features of zero variance (e.g., 1 across all samples)

preprocess(df,thr=0.9).shape

removing columns: {'fr_SH', 'NumAromaticCarbocycles', 'Chi0v', 'NOCount', 'SlogP_VSA5', 'VSA_EState2', 'Chi2n', 'NumHDonors', 'Chi1v', 'VSA_EState10', 'fr_NH2', 'Chi4n', 'Ipc', 'Chi3v', 'Chi1', 'NumHeteroatoms', 'Kappa1', 'Chi2v', 'RingCount', 'Chi3n', 'SMR_VSA1', 'Chi4v', 'VSA_EState6', 'SMR_VSA9', 'fr_Ar_N', 'NumAromaticRings', 'NumRotatableBonds', 'Chi0n'}

(25, 78)