Getting started with PLBenchmarks

from PLBenchmarks import targets
from IPython.core.display import HTML

Get the whole set of targets in the dataset

# it is initialized from the `PLBenchmarks/data/targets.yml` file
tgtset = targets.targetSet()
# to see which targets are available, one can get a list of names
tgtset.getNames()

['jnk1',
 'pde2',
 'thrombin',
 'p38',
 'ptp1b',
 'galectin',
 'cdk2',
 'cmet',
 'mcl1']

The targetSet is a dict, but can be converted to a pandas.DataFrame or a html string via targetSet.getDF(columns=None) or targetSet.getHTML(columns=None). The default None for columns means that all columns are printed. One can also define a subset of columns as a list:

HTML(tgtset.getHTML(columns=['name', 'fullname', 'pdb', 'references', 'numLigands', 'minDG', 'maxDG', 'associated_sets']))
name fullname pdb references numLigands minDG maxDG associated_sets
0 jnk1 c-Jun N-terminal kinase 1 2GMX [{'measurement': None}, {'calculation': ['10.1021/ja512751q', 'acs.jcim.9b00105']}] 21 -10.78111762414039 kcal/mol -7.353005882502171 kcal/mol [Schrodinger JACS]
1 pde2 phosphodiesterase 2 4D08,4D09,6EZF [{'measurement': ['10.1021/ml500262u', '10.1021/ja404449g']}, {'calculation': ['10.1038/s41598-018-23039-5']}] 21 -12.01124485346444 kcal/mol -8.812821938199054 kcal/mol None
2 thrombin thrombin 2ZFF [{'measurement': None}, {'calculation': ['10.1021/ja512751q', '10.1021/acs.jcim.9b00105']}] 11 -9.177820267686423 kcal/mol -7.480879541108987 kcal/mol [Schrodinger JACS]
3 p38 p38 alpha MAP kinase 1OUY, 3FLY [{'measurement': '10.1021/jm101423y'}, {'calculation': '10.1021/ja512751q, 10.1021/acs.jcim.9b00105'}] 34 -12.354423277849138 kcal/mol -8.546808630776018 kcal/mol [Schrodinger JACS]
4 ptp1b protein-tyrosine phosphatase 1B 2QBS [{'measurement': None}, {'calculation': '10.1021/ja512751q, 10.1021/acs.jcim.9b00105'}] 23 -12.584340587592953 kcal/mol -7.409826188396777 kcal/mol [Schrodinger JACS]
5 galectin galectin-3C 5E89,5E8A,5E88 [{'measurement': '10.1002/cbic.201600285'}, {'calculation': '10.1007/s10822-018-0110-5'}] 8 0.0 kcal/mol 0.0 kcal/mol None
6 cdk2 cyclin-dependent kinase 2 1H1Q, 2WEV [{'measurement': '10.1021/ci5004027'}, {'calculation': '10.1021/ja512751q, 10.1021/acs.jcim.9b00105'}] 16 -11.349056331748212 kcal/mol -7.09348579743778 kcal/mol [Schrodinger JACS]
7 cmet tyrosine-protein kinase Met; hepatocyte growth factor receptor (HGFR) 4R1Y [{'measurement': '10.1016/j.bmcl.2015.02.002'}, {'calculation': None}] 12 0.0 kcal/mol 0.0 kcal/mol [Merck KGaA FEP Benchmarks, YANK Benchmarks]
8 mcl1 Induced myeloid leukemia cell differentiation protein Mcl-1 4HW3 [{'measurement': None}, {'calculation': '10.1021/ja512751q, 10.1021/acs.jcim.9b00105'}] 42 0.0 kcal/mol 0.0 kcal/mol [Schrodinger JACS]

A target can be accessed with its name in two ways

jnk1 = tgtset['jnk1']
pde2 = tgtset.getTarget('pde2')

The target class

contains all the available information about one target of PLBenchmarks. It also has two member variables, _ligandSet and _edgeSet, which contain the information about the available ligand and edges of the respective target. A target can either be accessed from the targetSet (see cell before) or initialized using its name via

jnk1 = targets.target('jnk1')
# The data in the column is stored in a pandas.Series and can be accessed via
jnk1.getDF(columns=None)

id                                                                 1
name                                                            jnk1
fullname                                   c-Jun N-terminal kinase 1
netcharge                                                         xx
pdb                                                             2GMX
ligands            [lig_17124-1, lig_18624-1, lig_18625-1, lig_18...
references         [{'measurement': None}, {'calculation': ['10.1...
comments                                                        None
associated_sets                                   [Schrodinger JACS]
dtype: object

Access to the edgeSet and ligandSet in different formats is achieved by

jnk1_ligands = jnk1.getLigandSet()
jnk1_ligands_df = jnk1.getLigandSetDF()
HTML(jnk1.getLigandSetHTML(columns = ['name', 'ROMol', 'measurement', 'DerivedMeasurement']))
name measurement DerivedMeasurement measurement ROMol
ic50 doi comment e_ic50 dg e_dg doi_html
0 lig_17124-1 77 nM 10.1021/jm060199b table 1 cmpd 6t 38 nM -9.76481161912498 kcal/mol -10.185832695626143 kcal/mol Szczepankiewicz et al., J. Med. Chem. 2006 Mol dtype: object
1 lig_18624-1 570 nM 10.1021/jm060199b table 1 cmpd 6e 140 nM -8.571396114764765 kcal/mol -9.408403926601597 kcal/mol Szczepankiewicz et al., J. Med. Chem. 2006 Mol dtype: object
2 lig_18625-1 1100 nM 10.1021/jm060199b table 1 cmpd 6f 300 nM -8.179461879338154 kcal/mol -8.954045001030975 kcal/mol Szczepankiewicz et al., J. Med. Chem. 2006 Mol dtype: object
3 lig_18626-1 300 nM 10.1021/jm060199b table 1 cmpd 6g 70 nM -8.954045001030975 kcal/mol -9.821631925019588 kcal/mol Szczepankiewicz et al., J. Med. Chem. 2006 Mol dtype: object
4 lig_18627-1 580 nM 10.1021/jm060199b table 1 cmpd 6h 170 nM -8.56102781892315 kcal/mol -9.292655491812555 kcal/mol Szczepankiewicz et al., J. Med. Chem. 2006 Mol dtype: object
5 lig_18628-1 400 nM 10.1021/jm060199b table 1 cmpd 6i 120 nM -8.782539885935572 kcal/mol -9.500302701733787 kcal/mol Szczepankiewicz et al., J. Med. Chem. 2006 Mol dtype: object
6 lig_18629-1 420 nM 10.1021/jm060199b table 1 cmpd 6j 56 nM -8.753453044861022 kcal/mol -9.954661627304407 kcal/mol Szczepankiewicz et al., J. Med. Chem. 2006 Mol dtype: object
7 lig_18630-1 190 nM 10.1021/jm060199b table 1 cmpd 6k 4 nM -9.22634699650534 kcal/mol -11.527967281013156 kcal/mol Szczepankiewicz et al., J. Med. Chem. 2006 Mol dtype: object
8 lig_18631-1 120 nM 10.1021/jm060199b table 1 cmpd 6l 47 nM -9.500302701733787 kcal/mol -10.059111644635502 kcal/mol Szczepankiewicz et al., J. Med. Chem. 2006 Mol dtype: object
9 lig_18632-1 210 nM 10.1021/jm060199b table 1 cmpd 6m 95 nM -9.166681043279013 kcal/mol -9.639574994923331 kcal/mol Szczepankiewicz et al., J. Med. Chem. 2006 Mol dtype: object
10 lig_18633-1 180 nM 10.1021/jm060199b table 1 cmpd 6n 59 nM -9.258579818411201 kcal/mol -9.92355046515384 kcal/mol Szczepankiewicz et al., J. Med. Chem. 2006 Mol dtype: object
11 lig_18634-1 45 nM 10.1021/jm060199b table 1 cmpd 6o 3 nM -10.085035815247183 kcal/mol -11.699472396108563 kcal/mol Szczepankiewicz et al., J. Med. Chem. 2006 Mol dtype: object
12 lig_18635-1 4400 nM 10.1021/jm060199b table 1 cmpd 6p 670 nM -7.353005882502171 kcal/mol -8.475031685912471 kcal/mol Szczepankiewicz et al., J. Med. Chem. 2006 Mol dtype: object
13 lig_18636-1 3000 nM 10.1021/jm060199b table 1 cmpd 6q 1600 nM -7.581331303492182 kcal/mol -7.956083889099589 kcal/mol Szczepankiewicz et al., J. Med. Chem. 2006 Mol dtype: object
14 lig_18637-1 35 nM 10.1021/jm060199b table 1 cmpd 6r 16 nM -10.234859923437579 kcal/mol -10.701511284177176 kcal/mol Szczepankiewicz et al., J. Med. Chem. 2006 Mol dtype: object
15 lig_18638-1 38 nM 10.1021/jm060199b table 1 cmpd 6s 5 nM -10.185832695626143 kcal/mol -11.394937578728335 kcal/mol Szczepankiewicz et al., J. Med. Chem. 2006 Mol dtype: object
16 lig_18639-1 69 nM 10.1021/jm060199b table 1 cmpd 6u 14 nM -9.83020994328735 kcal/mol -10.78111762414039 kcal/mol Szczepankiewicz et al., J. Med. Chem. 2006 Mol dtype: object
17 lig_18652-1 14 nM 10.1021/jm060199b table 3 cmpd 18b 2 nM -10.78111762414039 kcal/mol -11.941195279431149 kcal/mol Szczepankiewicz et al., J. Med. Chem. 2006 Mol dtype: object
18 lig_18658-1 74 nM 10.1021/jm060199b table 3 cmpd 20a 11 nM -9.788503292300433 kcal/mol -10.924889274415738 kcal/mol Szczepankiewicz et al., J. Med. Chem. 2006 Mol dtype: object
19 lig_18659-1 110 nM 10.1021/jm060199b table 3 cmpd 20b 13 nM -9.552175576876945 kcal/mol -10.82529797985526 kcal/mol Szczepankiewicz et al., J. Med. Chem. 2006 Mol dtype: object
20 lig_18660-1 400 nM 10.1021/jm060199b table 3 cmpd 20c 45 nM -8.782539885935572 kcal/mol -10.085035815247183 kcal/mol Szczepankiewicz et al., J. Med. Chem. 2006 Mol dtype: object 
jnk1_edges = jnk1.getEdgeSet()
jnk1_edges_df = jnk1.getEdgeSetDF()
HTML(jnk1.getEdgeSetHTML())
0 1 Mol1 Mol2 exp. DeltaG [kcal/mol]
0 17124-1 18634-1 Mol Mol -0.32
1 18626-1 18624-1 Mol Mol 0.38
2 18636-1 18625-1 Mol Mol -0.60
3 18632-1 18624-1 Mol Mol 0.60
4 18635-1 18625-1 Mol Mol -0.83
5 18626-1 18658-1 Mol Mol -0.83
6 18639-1 18658-1 Mol Mol 0.04
7 18626-1 18625-1 Mol Mol 0.77
8 18638-1 18658-1 Mol Mol 0.40
9 18628-1 18624-1 Mol Mol 0.21
10 18631-1 18660-1 Mol Mol 0.72
11 18638-1 18634-1 Mol Mol 0.10
12 18626-1 18632-1 Mol Mol -0.21
13 18626-1 18630-1 Mol Mol -0.27
14 18631-1 18624-1 Mol Mol 0.93
15 18629-1 18627-1 Mol Mol 0.19
16 18634-1 18637-1 Mol Mol -0.15
17 18626-1 18627-1 Mol Mol 0.39
18 18631-1 18652-1 Mol Mol -1.28
19 18637-1 18631-1 Mol Mol 0.73
20 18626-1 18634-1 Mol Mol -1.13
21 18633-1 18624-1 Mol Mol 0.69
22 17124-1 18631-1 Mol Mol 0.26
23 18627-1 18630-1 Mol Mol -0.67
24 18659-1 18634-1 Mol Mol -0.53
25 18636-1 18624-1 Mol Mol -0.99
26 18626-1 18628-1 Mol Mol 0.17
27 18626-1 18660-1 Mol Mol 0.17
28 18626-1 18659-1 Mol Mol -0.60
29 18639-1 18634-1 Mol Mol -0.25
30 18635-1 18624-1 Mol Mol -1.22

Finally, the set out of ligands and edges can be visualized in a graph:

graph = jnk1.getGraph()

/opt/anaconda3/envs/off-demo/lib/python3.7/site-packages/networkx/drawing/nx_pylab.py:579: MatplotlibDeprecationWarning:
The iterable function was deprecated in Matplotlib 3.1 and will be removed in 3.3. Use np.iterable instead.
  if not cb.iterable(width):
_images/examples_graph.png

The ligandSet and ligand class

The ligandSet consists of a dict of ligands which are availabe for one target. It is accessible via target.getLigandSet(), but can also be initialized directly.

from PLBenchmarks import ligands

thrombin_ligands = ligands.ligandSet('thrombin')
HTML(thrombin_ligands.getHTML())
name smiles docked measurement DerivedMeasurement measurement ROMol
dg dh tds doi comment e_dg e_dh e_tds dg e_dg doi_html
0 lig_1a c1ccc(cc1)C[C@H](C(=O)N2CCC[C@H]2C(=O)NCc3cccc(c3)F)[NH3+] 03_docked/lig_1a/lig_1a.sdf -31.3 kJ/mol -13.1 kJ/mol -18.2 kJ/mol 10.1016/j.jmb.2009.04.051 Table 1 cmpd 1a 0.2 kJ/mol 0.9 kJ/mol 0.7 kJ/mol -7.480879541108987 kcal/mol 0.04780114722753346 kcal/mol Baum et al., Journal of Molecular Biology 2009 Mol dtype: object
1 lig_1b c1ccc(cc1)C[C@H](C(=O)N2CCC[C@H]2C(=O)NCc3cccc(c3)Cl)[NH3+] 03_docked/lig_1b/lig_1b.sdf -35.4 kJ/mol -37.1 kJ/mol 1.7 kJ/mol 10.1016/j.jmb.2009.04.051 Table 1 cmpd 1b 0.8 kJ/mol 1.1 kJ/mol 0.3 kJ/mol -8.460803059273422 kcal/mol 0.19120458891013384 kcal/mol Baum et al., Journal of Molecular Biology 2009 Mol dtype: object
2 lig_1c c1ccc(cc1)C[C@H](C(=O)N2CCC[C@H]2C(=O)NCc3cccc(c3)Br)[NH3+] 03_docked/lig_1c/lig_1c.sdf -35.8 kJ/mol -34.5 kJ/mol -1.3 kJ/mol 10.1016/j.jmb.2009.04.051 Table 1 cmpd 1c 0.7 kJ/mol 0.4 kJ/mol 0.3 kJ/mol -8.556405353728488 kcal/mol 0.1673040152963671 kcal/mol Baum et al., Journal of Molecular Biology 2009 Mol dtype: object
3 lig_1d c1ccc(cc1)C[C@H](C(=O)N2CCC[C@H]2C(=O)NCc3cccc(c3)I)[NH3+] 03_docked/lig_1d/lig_1d.sdf -34.5 kJ/mol -38.0 kJ/mol -3.5 kJ/mol 10.1016/j.jmb.2009.04.051 Table 1 cmpd 1d 0.3 kJ/mol 1.1 kJ/mol 0.8 kJ/mol -8.24569789674952 kcal/mol 0.07170172084130018 kcal/mol Baum et al., Journal of Molecular Biology 2009 Mol dtype: object
4 lig_3a Cc1cccc(c1)CNC(=O)[C@@H]2CCCN2C(=O)[C@@H](Cc3ccccc3)[NH3+] 03_docked/lig_3a/lig_3a.sdf -34.8 kJ/mol -28.5 kJ/mol -6.3 kJ/mol 10.1016/j.jmb.2009.04.051 Table 1 cmpd 3a 0.6 kJ/mol 0.7 kJ/mol 1.1 kJ/mol -8.317399617590821 kcal/mol 0.14340344168260036 kcal/mol Baum et al., Journal of Molecular Biology 2009 Mol dtype: object
5 lig_3b CCc1cccc(c1)CNC(=O)[C@@H]2CCCN2C(=O)[C@@H](Cc3ccccc3)[NH3+] 03_docked/lig_3b/lig_3b.sdf -32.9 kJ/mol -16.5 kJ/mol -16.4 kJ/mol 10.1016/j.jmb.2009.04.051 Table 1 cmpd 3b 0.5 kJ/mol 0.9 kJ/mol 0.4 kJ/mol -7.863288718929254 kcal/mol 0.11950286806883365 kcal/mol Baum et al., Journal of Molecular Biology 2009 Mol dtype: object
6 lig_5 c1ccc(cc1)C[C@H](C(=O)N2CCC[C@H]2C(=O)NCc3ccccc3)[NH3+] 03_docked/lig_5/lig_5.sdf -31.7 kJ/mol -13.6 kJ/mol -18.1 kJ/mol 10.1016/j.jmb.2009.04.051 Table 1 cmpd 5 0.0 kJ/mol 0.0 kJ/mol 0.0 kJ/mol -7.576481835564053 kcal/mol 0.0 kcal/mol Baum et al., Journal of Molecular Biology 2009 Mol dtype: object
7 lig_6a c1ccc(cc1)C[C@H](C(=O)N2CCC[C@H]2C(=O)NCc3cc(ccc3Cl)Cl)[NH3+] 03_docked/lig_6a/lig_6a.sdf -38.4 kJ/mol -41.3 kJ/mol 2.9 kJ/mol 10.1016/j.jmb.2009.04.051 Table 1 cmpd 6a 0.2 kJ/mol 0.4 kJ/mol 0.3 kJ/mol -9.177820267686423 kcal/mol 0.04780114722753346 kcal/mol Baum et al., Journal of Molecular Biology 2009 Mol dtype: object
8 lig_6b Cc1ccc(cc1CNC(=O)[C@@H]2CCCN2C(=O)[C@@H](Cc3ccccc3)[NH3+])Cl 03_docked/lig_6b/lig_6b.sdf -37.2 kJ/mol -33.5 kJ/mol -3.7 kJ/mol 10.1016/j.jmb.2009.04.051 Table 1 cmpd 6b 0.5 kJ/mol 1.7 kJ/mol 2.2 kJ/mol -8.891013384321225 kcal/mol 0.11950286806883365 kcal/mol Baum et al., Journal of Molecular Biology 2009 Mol dtype: object
9 lig_6e c1ccc(cc1)C[C@H](C(=O)N2CCC[C@H]2C(=O)NCc3cc(ccc3F)Cl)[NH3+] 03_docked/lig_6e/lig_6e.sdf -37.3 kJ/mol -41.0 kJ/mol 3.8 kJ/mol 10.1016/j.jmb.2009.04.051 Table 1 cmpd 6e 0.3 kJ/mol 2.1 kJ/mol 2.2 kJ/mol -8.914913957934989 kcal/mol 0.07170172084130018 kcal/mol Baum et al., Journal of Molecular Biology 2009 Mol dtype: object
10 lig_7a Cc1ccc(c(c1)CNC(=O)[C@@H]2CCCN2C(=O)[C@@H](Cc3ccccc3)[NH3+])C 03_docked/lig_7a/lig_7a.sdf -34.4 kJ/mol -31.9 kJ/mol -2.5 kJ/mol 10.1016/j.jmb.2009.04.051 Table 1 cmpd 7a 0.1 kJ/mol 1.3 kJ/mol 1.3 kJ/mol -8.221797323135755 kcal/mol 0.02390057361376673 kcal/mol Baum et al., Journal of Molecular Biology 2009 Mol dtype: object

The ligand classes can be accessed from the ligandSet by their name. Each ligand has information about experimental data, references, SMILES string and SDF file path of the docked structure. Additionally, there are functions to derive and process the primary data, which is then added to the pandas.Series as a new entry.

lig_6e = thrombin_ligands['lig_6e']
lig_1a = thrombin_ligands.getLigand('lig_6e')

The edgeSet and edge class

The edgeSet contains a dict of edges which are availabe for one target. It is accessible via target.getEdgeSet(), but can also be initialized directly.

from PLBenchmarks import edges

pde2_edges = edges.edgeSet('pde2')
HTML(pde2_edges.getHTML())
0 1 Mol1 Mol2 exp. DeltaG [kcal/mol]
0 49220392 49137530 Mol Mol 0.10
1 49932714 49137530 Mol Mol -1.30
2 49582468 49137530 Mol Mol -0.84
3 49396360 49137530 Mol Mol -0.85
4 50181001 49137530 Mol Mol -1.66
5 49585367 49137530 Mol Mol -1.56
6 49220392 49175828 Mol Mol -0.74
7 49220548 49220392 Mol Mol -0.03
8 49220548 49932129 Mol Mol 1.56
9 49582468 49932129 Mol Mol 0.66
10 49396360 49175828 Mol Mol -1.69
11 49175828 49580115 Mol Mol 1.89
12 49220548 49137374 Mol Mol -0.04
13 49220548 49580115 Mol Mol 1.13
14 49396360 49220548 Mol Mol -0.92
15 49932714 49582390 Mol Mol -0.91
16 49396360 49582390 Mol Mol -0.45
17 50181001 49582390 Mol Mol -1.26
18 50107616 49582390 Mol Mol -1.33
19 48168913 48271249 Mol Mol -0.45
20 49072088 48271249 Mol Mol -2.06
21 50107616 48271249 Mol Mol -1.43
22 49137374 48271249 Mol Mol 0.41
23 49932714 49175789 Mol Mol -1.25
24 49932714 49580115 Mol Mol -0.25
25 49932714 49582468 Mol Mol -0.47
26 48168913 49585367 Mol Mol 0.81
27 50107616 49585367 Mol Mol -0.16
28 48168913 48022468 Mol Mol -1.54
29 43249674 48022468 Mol Mol -0.19
30 48009208 43249674 Mol Mol -0.75
31 43249674 49175789 Mol Mol 0.65
32 49175789 49072088 Mol Mol 2.31
33 48009208 49137374 Mol Mol -0.27 
pde2_edges.keys()

dict_keys(['edge_49220392_49137530', 'edge_49932714_49137530', 'edge_49582468_49137530', 'edge_49396360_49137530', 'edge_50181001_49137530', 'edge_49585367_49137530', 'edge_49220392_49175828', 'edge_49220548_49220392', 'edge_49220548_49932129', 'edge_49582468_49932129', 'edge_49396360_49175828', 'edge_49175828_49580115', 'edge_49220548_49137374', 'edge_49220548_49580115', 'edge_49396360_49220548', 'edge_49932714_49582390', 'edge_49396360_49582390', 'edge_50181001_49582390', 'edge_50107616_49582390', 'edge_48168913_48271249', 'edge_49072088_48271249', 'edge_50107616_48271249', 'edge_49137374_48271249', 'edge_49932714_49175789', 'edge_49932714_49580115', 'edge_49932714_49582468', 'edge_48168913_49585367', 'edge_50107616_49585367', 'edge_48168913_48022468', 'edge_43249674_48022468', 'edge_48009208_43249674', 'edge_43249674_49175789', 'edge_49175789_49072088', 'edge_48009208_49137374'])

The edge classes can be accessed from the edgeSet by their name. They are lightweight and provide only access to a pandas.DataFrame and a dict:

edge_49220392_49137530 = pde2_edges.getEdge('edge_49220392_49137530')
df = edge_49220392_49137530.getDF()
edge_49220392_49137530.getDict()

{'edge_49220392_49137530': ['lig_49220392', 'lig_49137530']}