models.py 18.4 KB
Newer Older
Hervé  MENAGER's avatar
Hervé MENAGER committed
1
2
3
4
from __future__ import unicode_literals

from django.db import models

5
from .ws import get_pubmed_info, get_epo_info, get_uniprot_info, get_taxonomy_info, get_go_info, get_pfam_info
6

Hervé  MENAGER's avatar
Hervé MENAGER committed
7
8
9
10
11
12
13
14
15
16
17
class Bibliography(models.Model):
    """
    Bibliography data table
    """
    SOURCES = (
        ('PM', 'PubMed article'),
        ('PT', 'Patent')
    )
    source = models.CharField('Bibliographic type', max_length=2, choices=SOURCES)
    id_source = models.CharField('Bibliographic ID', max_length=25)
    title = models.CharField('Title', max_length=300)
18
    journal_name = models.CharField('Journal name', max_length=50, null=True)
Hervé  MENAGER's avatar
Hervé MENAGER committed
19
20
    authors_list = models.CharField('Authors list', max_length=500)
    biblio_year = models.PositiveSmallIntegerField('Year')
21
22
23
24
25
26
27
    cytotox = models.BooleanField('Cytotoxicity data', default=False)
    in_silico = models.BooleanField('in silico study performed', default=False)
    in_vitro = models.BooleanField('in vitro study performed', default=False)
    in_vivo = models.BooleanField('in vivo study performed', default=False)
    in_cellulo = models.BooleanField('in cellulo study performed', default=False)
    pharmacokinetic = models.BooleanField('pharmacokinetic study performed', default=False)
    xray = models.BooleanField('contains xray data', default=False)
Hervé  MENAGER's avatar
Hervé MENAGER committed
28
29

    def save(self, *args, **kwargs):
30
31
32
33
34
35
36
37
        if self.source == 'PM':
            info = get_pubmed_info(self.id_source)
        else:
            info = get_epo_info(self.id_source)
        self.title = info['title']
        self.journal_name = info['journal_name']
        self.authors_list = info['authors_list']
        self.biblio_year = info['biblio_year']
Hervé  MENAGER's avatar
Hervé MENAGER committed
38
39
        super(Bibliography, self).save(*args, **kwargs)

Hervé  MENAGER's avatar
Hervé MENAGER committed
40
41
42
    class Meta:
        verbose_name_plural = "bibliographies"

43

Hervé  MENAGER's avatar
Hervé MENAGER committed
44
45

class Taxonomy(models.Model):
46
    taxonomy_id = models.DecimalField('NCBI TaxID', unique=True, max_digits=9, decimal_places=0)
Hervé  MENAGER's avatar
Hervé MENAGER committed
47
    name = models.CharField('Organism name', max_length=200)
48
49
50
51
52
53

    def save(self, *args, **kwargs):
        info = get_taxonomy_info(self.taxonomy_id)
        self.name = info['scientific_name']
        super(Taxonomy, self).save(*args, **kwargs)

54
55
56
    def __str__(self):
        return self.name

Hervé  MENAGER's avatar
Hervé MENAGER committed
57
58
    class Meta:
        verbose_name_plural = "taxonomies"
Hervé  MENAGER's avatar
Hervé MENAGER committed
59
60
61
62
63

class MolecularFunction(models.Model):
    go_id = models.CharField('Gene Ontology ID', unique=True, max_length=10) # GO term id format: 'GO:0000000' 
    description = models.CharField('description', max_length=500)

64
65
66
67
68
    def save(self, *args, **kwargs):
        info = get_go_info(self.go_id)
        self.description = info['label']
        super(MolecularFunction, self).save(*args, **kwargs)

69
70
71
    def __str__(self):
        return self.description

Hervé  MENAGER's avatar
Hervé MENAGER committed
72
73
74
75
76
77
78
79
80
class Protein(models.Model):
    uniprot_id = models.CharField('Uniprot ID', unique=True, max_length=10)
    recommended_name_long = models.CharField('Uniprot Recommended Name (long)', max_length=75)
    short_name = models.CharField('Short name', max_length=50)
    gene_name = models.CharField('Gene name', unique=True, max_length=30)
    entry_name = models.CharField('Entry name', max_length=30)
    organism = models.ForeignKey('Taxonomy')
    molecular_functions = models.ManyToManyField(MolecularFunction)

81
    def save(self, *args, **kwargs):
82
        info = get_uniprot_info(self.uniprot_id)
83
        self.recommended_name_long = info['recommended_name']
84
85
86
87
88
89
90
91
92
        self.gene_name = info['gene']
        self.entry_name = info['entry_name']
        try:
            taxonomy = Taxonomy.objects.get(taxonomy_id=info['organism'])
        except Taxonomy.DoesNotExist:
            taxonomy = Taxonomy()
            taxonomy.taxonomy_id = info['organism']
            taxonomy.save()
        self.organism = taxonomy
93
        super(Protein, self).save(*args, **kwargs)
94
95
96
97
98
99
100
101
        for go_id in info['molecular_functions']:
            try:
                mol_function = MolecularFunction.objects.get(go_id=go_id)
            except MolecularFunction.DoesNotExist:
                mol_function = MolecularFunction()
                mol_function.go_id = go_id
                mol_function.save()
            self.molecular_functions.add(mol_function)
102

103
104
105
    def __str__(self):
        return '{} ({})'.format(self.uniprot_id, self.recommended_name_long)

Hervé  MENAGER's avatar
Hervé MENAGER committed
106
107
108
109
class Domain(models.Model):
    pfam_acc = models.CharField('Pfam Accession', max_length=10, unique=True) 
    pfam_id = models.CharField('Pfam Family Identifier', max_length=20) 
    pfam_description = models.CharField('Pfam Description', max_length=100)
110
111
112
113
114
115
116
    domain_family = models.CharField('Domain family', max_length=25)  #TODO: what is this field? check database contents

    def save(self, *args, **kwargs):
        info = get_pfam_info(self.pfam_acc)
        self.pfam_id = info['id']
        self.pfam_description = info['description']
        super(Domain, self).save(*args, **kwargs)
Hervé  MENAGER's avatar
Hervé MENAGER committed
117

118
119
120
    def __str__(self):
        return '{} ({}-{})'.format(self.pfam_acc, self.pfam_id, self.pfam_description)

121
class ProteinDomainComplex(models.Model):
122
123
    protein = models.ForeignKey('Protein')
    domain = models.ForeignKey('Domain')
124
125
    ppc_copy_nb = models.IntegerField('Number of copies of the protein in the complex')
    
Hervé  MENAGER's avatar
Hervé MENAGER committed
126
127
    class Meta:
        verbose_name_plural = "complexes"
128

129
130
131
    def __str__(self):
        return '{}-{}'.format(self.protein_id, self.domain_id)

132
133
134
135
136
class ProteinDomainBoundComplex(ProteinDomainComplex):
    ppp_copy_nb_per_p = models.IntegerField('Number of copies of the protein in the pocket')
    pockets_nb = models.IntegerField('Total number of pockets in the complex')
    class Meta:
        verbose_name_plural = "bound complexes"
Hervé  MENAGER's avatar
Hervé MENAGER committed
137
    
138
139
140
class ProteinDomainPartnerComplex(ProteinDomainComplex):
    class Meta:
        verbose_name_plural = "partner complexes"
Hervé  MENAGER's avatar
Hervé MENAGER committed
141

142
143
144
class Symmetry(models.Model):
    code = models.CharField('Symmetry code', max_length=2)
    description = models.CharField('Description', max_length=300)
Hervé  MENAGER's avatar
Hervé MENAGER committed
145

146
147
148
    class Meta:
        verbose_name_plural = "symmetries"

149
150
151
    def __str__(self):
        return '{} ({})'.format(self.code, self.description)

152
153
154
155
156
157
158

class Disease(models.Model):
    name = models.CharField('Disease', max_length=30, unique=True) # is there any database/nomenclature for diseases?

    def __str__(self):
        return self.name

159
class Ppi(models.Model):
160
    pdb_id = models.CharField('PDB ID', max_length=4, null=True)
161
    symmetry = models.ForeignKey(Symmetry)
162
    diseases = models.ManyToManyField(Disease)
Hervé  MENAGER's avatar
Hervé MENAGER committed
163

164
165
166
    def __str__(self):
        return '{} PPI, PDB:{}'.format(self.symmetry.description, self.pdb_id or 'unknown')

167
168
169
170
171
172
173
174
175
176
177
178
179
    def get_ppi_complexes(self):
        """
        return all ppi complexes belonging to this ppi
        """
        return PpiComplex.objects.filter(ppi=self)

    def get_ppi_bound_complexes(self):
        """
        return bound ppi complexes belonging to this ppi
        """
        #this is the less efficient query ever seen, FIXME
        return PpiComplex.objects.filter(ppi=self, complex__in=ProteinDomainBoundComplex.objects.all())

180

Hervé  MENAGER's avatar
Hervé MENAGER committed
181
182
183
184
185
186
187
188
class PpiComplex(models.Model):
    ppi = models.ForeignKey(Ppi)
    complex = models.ForeignKey(ProteinDomainComplex)
    cc_nb = models.IntegerField('Number of copies of the complex in the PPI', default=1)

    class Meta:
        verbose_name_plural = "Ppi complexes"

189
190
191
    def __str__(self):
        return 'PPI {}, Complex {} ({})'.format(self.ppi, self.complex, self.cc_nb)

Hervé  MENAGER's avatar
Hervé MENAGER committed
192

193
class Compound(models.Model):
Hervé  MENAGER's avatar
Hervé MENAGER committed
194
    canonical_smile = models.CharField('Canonical Smile', unique=True, max_length=500)
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
    #TODO index this table on canonical_smile
    is_macrocycle = models.BooleanField('Contains one or more macrocycles') 
    aromatic_ratio = models.DecimalField('Aromatic ratio', max_digits=3, decimal_places=2)
    balaban_index = models.DecimalField('Balaban index', max_digits=3, decimal_places=2)  
    fsp3 = models.DecimalField('Fsp3', max_digits=3, decimal_places=2)
    gc_molar_refractivity = models.DecimalField('GC Molar Refractivity', max_digits=5, decimal_places=2)
    log_d = models.DecimalField('LogD (Partition coefficient octanol-1/water, with pKa information)', max_digits=4, decimal_places=2)
    a_log_p = models.DecimalField('ALogP (Partition coefficient octanol-1/water)', max_digits=4, decimal_places=2)
    mean_atom_vol_vdw = models.DecimalField('Mean atom volume computed with VdW radii', max_digits=4, decimal_places=2)
    molecular_weight = models.DecimalField('Molecular weight', max_digits=6, decimal_places=2)
    nb_acceptor_h = models.IntegerField('Number of hydrogen bond acceptors')
    nb_aliphatic_amines = models.IntegerField('Number of aliphatics amines')
    nb_aromatic_bonds = models.IntegerField('Number of aromatic bonds')
    nb_aromatic_ether = models.IntegerField('Number of aromatic ethers')
    nb_aromatic_sssr = models.IntegerField('Number of aromatic Smallest Set of System Rings (SSSR)')
    nb_atom = models.IntegerField('Number of atoms') 
    nb_atom_non_h = models.IntegerField('Number of non hydrogen atoms') 
    nb_benzene_like_rings = models.IntegerField('Number of benzene-like rings')
    nb_bonds = models.IntegerField('Number of bonds')
    nb_bonds_non_h = models.IntegerField('Number of bonds not involving a hydrogen')  
    nb_br = models.IntegerField('Number of Bromine atoms')  
    nb_c = models.IntegerField('Number of Carbon atoms')  
    nb_chiral_centers = models.IntegerField('Number of chiral centers')  
    nb_circuits = models.IntegerField('Number of circuits')  
    nb_cl = models.IntegerField('Number of Chlorine atoms')  
    nb_csp2 = models.IntegerField('Number of sp2-hybridized carbon atoms')  
    nb_csp3 = models.IntegerField('Number of sp3-hybridized carbon atoms')  
    nb_donor_h = models.IntegerField('Number of hydrogen bond donors')  
    nb_double_bonds = models.IntegerField('Number of double bonds')  
    nb_f = models.IntegerField('Number of fluorine atoms')  
    nb_i = models.IntegerField('Number of iodine atoms')  
    nb_multiple_bonds = models.IntegerField('Number of multiple bonds')  
    nb_n = models.IntegerField('Number of nitrogen atoms')
    nb_o = models.IntegerField('Number of oxygen atoms')  
    nb_rings = models.IntegerField('Number of rings')  
    nb_rotatable_bonds = models.IntegerField('Number of rotatable bonds')  
    randic_index = models.DecimalField('Randic index', max_digits=4, decimal_places=2)  
    rdf070m = models.DecimalField('RDF070m, radial distribution function weighted by the atomic masses at 7Å', max_digits=5, decimal_places=2)  
    rotatable_bond_fraction = models.DecimalField('Fraction of rotatable bonds', max_digits=3, decimal_places=2)  
    sum_atom_polar = models.DecimalField('Sum of atomic polarizabilities', max_digits=5, decimal_places=2)  
    sum_atom_vol_vdw = models.DecimalField('Sum of atom volumes computed with VdW radii', max_digits=6, decimal_places=2)  
    tpsa = models.DecimalField('Topological Polar Surface Area (TPSA)', max_digits=5, decimal_places=2)  
    ui = models.DecimalField('Unsaturation index', max_digits=4, decimal_places=2)  
    wiener_index = models.IntegerField('Wiener index')  
    common_name = models.CharField('Common name', unique=True, max_length=20, blank=True, null=True)  
    pubchem_id = models.CharField('Pubchem ID', unique=True, max_length=10, blank=True, null=True)  
    chemspider_id = models.CharField('Chemspider ID', unique=True, max_length=10, blank=True, null=True)  
    chembl_id = models.CharField('Chembl ID', unique=True, max_length=30, blank=True, null=True)  
    iupac_name = models.CharField('IUPAC name', unique=True, max_length=255, blank=True, null=True)  
244
    mddr_compound = models.ForeignKey('MDDRCompoundImport', blank=True, null=True)  
245
246
247
248
249
250
251
252
253


class MDDRCompoundImport(models.Model):
    mddr_compound_id = models.IntegerField('MDDR compound ID')  
    mddr_name = models.CharField('MDDR name', max_length=40)  
    dvpmt_phase = models.CharField('Development phase', max_length=20)  
    canonical_smile = models.CharField('Canonical Smile', max_length=500, unique=True, blank=True, null=True)  
    #TODO index this table on canonical_smile
    db_import_date = models.DecimalField('MDDR release year/month', max_digits=6, decimal_places=0)  
Hervé  MENAGER's avatar
Hervé MENAGER committed
254

255
256
257
258
    class Meta:
        # over multiple releases of the MDDR database, the same compound can evolve in its development phase
        # the same compound can have different names and development phases in the same MDDR release
        unique_together = (('mddr_compound_id', 'mddr_name', 'dvpmt_phase'),)
Hervé  MENAGER's avatar
Hervé MENAGER committed
259
        verbose_name_plural = "MDDR compound imports"
Hervé  MENAGER's avatar
Hervé MENAGER committed
260
261


262
class MDDRCompoundActivityClass(models.Model):
263
    mddr_compound = models.ForeignKey(MDDRCompoundImport)  
264
    activity_class = models.CharField('Activity Class', max_length=100)  
Hervé  MENAGER's avatar
Hervé MENAGER committed
265
266

    class Meta:
267
        unique_together = (('mddr_compound', 'activity_class'),)
Hervé  MENAGER's avatar
Hervé MENAGER committed
268
        verbose_name_plural = "MDDR compound activity classes"
Hervé  MENAGER's avatar
Hervé MENAGER committed
269
270


271
272
273
274
class MDDRSimilarity(models.Model):
    canonical_smile_ippidb = models.CharField('Canonical Smile for IPPIDB compound', max_length=500, unique=True, blank=True, null=True)
    canonical_smile_mddr = models.CharField('Canonical Smile for MDDR Compound', max_length=500, unique=True, blank=True, null=True)
    tanimoto = models.DecimalField('Tanimoto', max_digits=6, decimal_places=5)  
Hervé  MENAGER's avatar
Hervé MENAGER committed
275
276

    class Meta:
277
        unique_together = (('canonical_smile_ippidb', 'canonical_smile_mddr'),)
Hervé  MENAGER's avatar
Hervé MENAGER committed
278
        verbose_name_plural = "MDDR similarities"
279

280
281
class CellLine(models.Model):
    name = models.CharField('Name', max_length=50, unique=True)
Hervé  MENAGER's avatar
Hervé MENAGER committed
282

283
284
285
286
287
288
289
290
291
292
class TestActivityDescription(models.Model):
    TEST_TYPES = (
        ('BIOCH', 'Biochemical assay'),
        ('CELL', 'Cellular assay')
    )
    TEST_MODULATION_TYPES = (
        ('B', 'Binding'),
        ('I', 'Inhibition'),
        ('S', 'Stabilization')
    )
293
294
    biblio = models.ForeignKey(Bibliography)
    ppi = models.ForeignKey(Ppi, blank=True, null=True)  
295
296
297
298
299
300
    test_name = models.CharField('Test name', max_length=100)  
    test_type = models.CharField('Test type', max_length=5, choices=TEST_TYPES)
    test_modulation_type = models.CharField('Test modulation type', max_length=1, choices=TEST_MODULATION_TYPES)
    nb_active_compounds = models.IntegerField('Total number of active compounds')  
    cell_line = models.ForeignKey(CellLine)

Hervé  MENAGER's avatar
Hervé MENAGER committed
301
    def get_complexes(self):
302
303
304
305
306
307
308
309
        """
        get the complexes tested for this PPI
        depends on the modulation type
        """
        if self.test_modulation_type=='I':
            return self.ppi.get_ppi_complexes()
        else:
            return self.ppi.get_ppi_bound_complexes()
310
311

class CompoundActivityResult(models.Model):
312
313
314
315
316
    MODULATION_TYPES = (
        ('B', 'Binding'),
        ('I', 'Inhibition'),
        ('S', 'Stabilization')
    )
Hervé  MENAGER's avatar
Hervé MENAGER committed
317
318
319
320
321
322
    ACTIVITY_TYPES = (
        ('pIC50','pIC50 (half maximal inhibitory concentration, -log10)'),
        ('pEC50','pEC50 (half maximal effective concentration, -log10)'),
        ('pKd','pKd (dissociation constant, -log10)'),
        ('pKi','pKi (inhibition constant, -log10)'),
    )
323
324
    compound = models.ForeignKey(Compound)  
    test_activity_description = models.ForeignKey(TestActivityDescription)  
Hervé  MENAGER's avatar
Hervé MENAGER committed
325
    activity_type = models.CharField('Activity type', max_length=5, choices=ACTIVITY_TYPES)
326
    activity = models.DecimalField('Activity', max_digits=12, decimal_places=10)  
327
    modulation_type = models.CharField('Modulation type', max_length=1, choices=MODULATION_TYPES)  
Hervé  MENAGER's avatar
Hervé MENAGER committed
328
329

    class Meta:
330
        unique_together = (('compound', 'test_activity_description', 'activity_type'),)
Hervé  MENAGER's avatar
Hervé MENAGER committed
331

332
class TestCytotoxDescription(models.Model):
333
    biblio = models.ForeignKey(Bibliography)  
334
335
336
    test_name = models.CharField('Cytotoxicity test name', max_length=100)  
    cell_line = models.ForeignKey(CellLine)
    compound_concentration = models.DecimalField('Compound concentration in μM', max_digits=7, decimal_places=3, blank=True, null=True)  
Hervé  MENAGER's avatar
Hervé MENAGER committed
337

338
class CompoundCytotoxicityResult(models.Model):
339
340
    compound = models.ForeignKey(Compound)  
    test_cytotoxicity_description = models.ForeignKey(TestCytotoxDescription)
341
    toxicity = models.BooleanField('Toxicity', default=False)
Hervé  MENAGER's avatar
Hervé MENAGER committed
342
343

    class Meta:
344
        unique_together = (('compound', 'test_cytotoxicity_description'),)
345
346
347
348
349

class AdministrationMode(models.Model):
    name = models.CharField('Administration mode', max_length=20, blank=True, null=True)

class TestPKDescription(models.Model):
350
    biblio = models.ForeignKey(Bibliography)  
351
352
353
354
355
356
357
    test_name = models.CharField('Pharmacokinetic test name', max_length=100)  
    organism = models.ForeignKey(Taxonomy)
    administration_mode = models.ForeignKey(AdministrationMode, blank=True, null=True)  
    dose = models.DecimalField('Dose in mg/kg', max_digits=7, decimal_places=4, blank=True, null=True)  
    dose_interval = models.IntegerField('Dose interval, in hours', blank=True, null=True)  

class CompoundPKResult(models.Model):
358
359
    compound = models.ForeignKey(Compound)  
    test_pk_description = models.ForeignKey(TestPKDescription)
360
361
362
363
364
365
366
367
    tolerated = models.NullBooleanField('Tolerated', null=True)
    auc = models.IntegerField('Area under curve (ng.mL-1.hr)', blank=True, null=True)  
    clearance = models.DecimalField('Clearance (mL/hr)', max_digits=7, decimal_places=3, blank=True, null=True)  
    cmax = models.DecimalField('Maximal concentration (ng/mL)', max_digits=7, decimal_places=3, blank=True, null=True)  
    oral_bioavailability = models.IntegerField('Oral Bioavailability (%F)', blank=True, null=True)  
    t_demi = models.IntegerField('t½', blank=True, null=True)  
    t_max = models.IntegerField('tmax', blank=True, null=True)  
    voldistribution = models.DecimalField('Volume distribution (Vd)', max_digits=5, decimal_places=2, blank=True, null=True)  
Hervé  MENAGER's avatar
Hervé MENAGER committed
368
369

    class Meta:
370
        unique_together = (('compound', 'test_pk_description'),)
Hervé  MENAGER's avatar
Hervé MENAGER committed
371
372


373
class CompoundAction(models.Model):
374
375
    ACTIVATION_MODES = (
        ('O', 'Orthosteric'),
376
        ('A', 'Allosteric')
377
    )
378
    compound = models.ForeignKey(Compound)  
379
    activation_mode = models.CharField('Activation mode', max_length=1, choices=ACTIVATION_MODES)  
380
    ppi = models.ForeignKey(Ppi)
381
382
    pdb_id = models.CharField('PDB ID', max_length=4)
    nb_copy_compounds = models.IntegerField('Number of copies for the compound')  
Hervé  MENAGER's avatar
Hervé MENAGER committed
383
384

    class Meta:
Hervé  MENAGER's avatar
Hervé MENAGER committed
385
386
387
        unique_together = (('ppi', 'compound', 'activation_mode', 'pdb_id'),)

    def get_complexes(self):
388
389
390
391
392
        """
        get the complexes involved in the compound action
        which are always the bound complexes
        """
        return ppi.get_ppi_bound_complexes()
Hervé  MENAGER's avatar
Hervé MENAGER committed
393

394
class RefCompoundBiblio(models.Model):
395
396
    compound = models.ForeignKey(Compound)  
    bibliography = models.ForeignKey(Bibliography)  
397
    compound_name = models.CharField('Compound name in the publication', max_length=50)  
Hervé  MENAGER's avatar
Hervé MENAGER committed
398
399

    class Meta:
400
        unique_together = (('compound', 'bibliography'),)