Skip to content

Tutorial 2.5_Database

Jump to bottom
LucaChiesa edited this page Nov 7, 2024 · 1 revision

Introduction

In this tutorial you will learn how to set up the .yaml filter file which queries any of the annotated structure databases supported by LIT-AlphaFold.

  1. GPCRdb for receptors
  2. GPCRdb for G-protein
  3. KLIFS
  4. ChannelsDB
  5. RCSB

GPCRdb for receptors

GPCRdb contains structure annotation for G-protein coupled receptor structures. The filter file below contains all the implemented fields used for template filtering. Certain fields accept only specific values, which are indicated in the example if possible.

'database': 'GPCRdb_r'
'excluded_pdb': [] # Common filed to all filter files
'subset_pdb': []   # Common filed to all filter files

# list fields

'signalling_protein': ['G protein', 'Arrestin']
'protein':
   - 'cxcr4_human' # All protins have their UniProt name in lowercase letters followed by the species
'excluded_protein':
   - 'ccr5_human' # All protins have their UniProt name in lowercase letters followed by the species
'ligand_function':
   - 'Agonist'
   - 'Agonist (partial)'
   - 'Partial agonist'
   - 'Allosteric agonist'
   - 'Antagonist'
   - 'Allosteric antagonist'
   - 'Inverse agonist'
   - 'Ago-PAM' #Agonist and positive allosteric modulator
   - 'PAM' #Positive allosteric modulator
   - 'NAM' #Negative allosteric modulator
   - 'unknown'
'class': 
   - 'Class A (Rhodopsin)'
   - 'Class B1 (Secretin)'
   - 'Class B2 (Adhesion)'
   - 'Class C (Glutamate)'
   - 'Class D1 (Ste2-like fungal pheromone)'
   - 'Class F (Frizzled)'
   - 'Class T (Taste 2)'
'species':
   - 'Homo sapiens'
   - 'Xenopus laevis' #African clawed frog
   - 'Bos taurus' #Bovine
   - 'Canis lupus familiaris' #Dog
   - 'Cavia porcellus' #Guinea-pig
   - 'Hasarius adansoni'
   - 'Todarodes pacificus' #Japanese flying squid
   - 'Mus musculus' #Mouse
   - 'Oryctolagus cuniculus' #Rabbit
   - 'Rattus norvegicus' #Rat
   - 'Human cytomegalovirus'
   - 'Saccharomyces cerevisiae'
   - 'Epstein-Barr virus'
   - 'Meleagris gallopavo' #Wild turkey
   - 'Danio rerio' #Zebrafish
'state': ['Active', 'Inactive', 'Intermediate', 'Other']
'type': ['Electron microscopy', 'X-ray diffraction']

# boolean fields

'apo': False # True

# single values fields

'publication_date': '2100-01-01' #YYYY-MM-DD
'resolution': 10.0

The fields are divided in three categories depending on the accepted datatype:

  • list, these fields have a list of acceptable values indicated between square brackets [] or by bullet points with -.
  • boolean, only the apo field, selects apo-protein structures if True or holo-protein structures if False, if not included selects both
  • single value, used for publication_date and resoultion, selects entries with a lower value than the specified one.

In the case of CXCR4 we might be interested in removing all templates of CXCR4, and selecting as templates only human receptors, bound to an antagonist, with resolution less than 3.5 and pusblished before 2020-03-08. The filter file in this case will be:

'database': 'GPCRdb_r'

# list fields

'excluded_protein':
   - 'cxcr4_human' # All protins have their UniProt name in lowercase letters followed by the species
'ligand_function':
   - 'Antagonist'
'species':
   - 'Homo sapiens'

# single values fields

'publication_date': '2020-03-08'
'resolution': 3.5

GPCRdb for G-protein

GPCRdb contains relevant information regarding the structures of G-protein bound to a GPCR. Template structures can be filtered to select only structures of G-protein bound to a receptor, and based on their species, and G-protein subtype.

'database': 'GPCRdb_r'
'excluded_pdb': [] # Common filed to all filter files
'subset_pdb': []   # Common filed to all filter files

# list fields

'species':
  - 'human'
  - 'rat'
  - 'bovin'

'excluded_protein':
  - 'gnai1'
  - 'gbb1'
  - 'gbg2'

'subset_protein':
  - 'gnai1'
  - 'gbb1'
  - 'gbg2'

# single values fields

'publication_date': '2100-01-01' #YYYY-MM-DD
'resolution': 10.0

The fields follow a similar syntax as the one for GPCRs.

KLIFS

KLIFS is a kinase database containing information on catalytic kinase domains. By querying this database it is possible to select structures based on the conformation of the DFG motif, or of the αC-helix, as well as based on the presence, or absence, of the salt bridge between KIII.17 and EαC.24.

'database': 'KLIFS'
'excluded_pdb': [] # Common filed to all filter files
'subset_pdb': []   # Common filed to all filter files

'DFG':
 - 'out'
 - 'out-like'
 - 'in'

'ac_helix':
  - 'out'
  - 'in'

'salt_bridge_17_24': True #False

The DFG and ac_helix fields take as input a list of acceptable conformations, while the salt_bridge_17_24 field selects structure that have (True) or do not have (False) a salt bridge.

ChannelsDB

ChannelsDB contains information regarding channels, pores and tunnels found in biomacromolecules.

Channel information has been extracted using different parameters of the MOLE algorithm, therefore queries are adjusted to handle possible different values depending on the used approach. The found channels were divided in multiple categories, allowing the user to select only a subset of these entries depending on their query.

For each channel property it is possible to give an exact value that must coincide with the query, or it is possible to set a maximum (max_) or minium (min_) threshold for the property.

'database': 'ChannelsDB'
'excluded_pdb': [] # Common filed to all filter files
'subset_pdb': []   # Common filed to all filter files

'ReviewedChannels':
  'Type':   #See https://channelsdb2.biodata.ceitec.cz/documentation.html#db-nomenclature for more information on channel nomenclature
    - 'Channel'
    - 'Tunnel'
    - 'Pore'
    - 'Solvent tunnel'
    - 'Substrate tunnel'
    - 'Substrate/Product tunnel'
    - 'Product tunnel'
    - 'Water channel'
    - 'Ion channel'
    - 'Hydrophobic channel'
    - 'Peptide channel'
    - 'Nucleotide channel'
  'Charge': 0
  'min_Charge': 0
  'max_Charge': 0
  'NumPositives': 0
  'NumNegatives': 0
  'Hydrophobicity': 0
  'Hydropathy': 0
  'Polarity': 0
  'Mutability': 0

'CSATunnels'
'CofactorTunnels'
'CognateTunnels'

RCSB

Information regarding the templates can be extracted directly from the Protein Data Bank (PDB).

While explicit information regarding the protein structure are not available, it is possible to select structures based on the presence or absence of specific ligands, based on their HET code.

Additionally, templates could be selected based on different types of protein classification schemes (CATH, SCOP2B,ECOD).

'database': 'RCSB'
'excluded_pdb': [] # Common filed to all filter files
'subset_pdb': []   # Common filed to all filter files

'apo': False #True
'holo': False #True

# Select/exclude structures where a ligand with the given HET code is present
'HET':
  - "HOH"
'exclude_HET':
  - "HOH"

# Select/exclude structure based on the evolutionary classification for the protein
'protein_class':
  'source':
    - 'CATH'
    - 'SCOP2B'
    - 'ECOD'
  'name':
    - 'Rhodopsin 7-helix transmembrane proteins' #see https://www.cathdb.info/browse/tree
    - 'G protein-coupled receptor-like' #see https://scop.berkeley.edu/ver=2.08
    - '7tm_1_1' #see http://prodata.swmed.edu/ecod/complete/

'excluded_protein_class':
  'source':
    - 'CATH'
  'name':
    - 'Rhodopsin 7-helix transmembrane proteins'

Conclusion

In this tutorial you have learned how to setup a query file for filtering template structures using Database annotations.

Clone this wiki locally