OmniPath table ingestion¶

omnipath_client.to_annnet builds a graph from an OmniPath-style interaction table. AnnNet reaches no knowledge base of its own: the client for a knowledge base is what returns AnnNet objects. This notebook uses a local table so it is deterministic.

In [1]:

import annnet as an
import omnipath_client as oc

an.info()

import annnet as an import omnipath_client as oc an.info()

---------------------------------------------------------------------------
ModuleNotFoundError                       Traceback (most recent call last)
Cell In[1], line 2
      1 import annnet as an
----> 2 import omnipath_client as oc
      4 an.info()

ModuleNotFoundError: No module named 'omnipath_client'

Build from a local OmniPath-style table¶

In [2]:

import polars as pl

interactions = pl.DataFrame(
    {
        'source': ['EGF', 'EGFR', 'EGFR', 'EGFR', 'RAS', 'MEK'],
        'target': ['EGFR', 'RAS', 'RAS', 'GRB2', 'MEK', 'ERK'],
        'interaction_id': [
            'EGF_EGFR',
            'EGFR_RAS_primary',
            'EGFR_RAS_secondary',
            'EGFR_GRB2_complex',
            'RAS_MEK',
            'MEK_ERK',
        ],
        'is_directed': [True, True, True, False, True, True],
        'curation_score': [0.95, 0.88, 0.63, 0.76, 0.82, 0.79],
        'consensus_direction': [1, 1, 1, 0, 1, 1],
        'source_database': [
            'omnipath',
            'omnipath',
            'literature',
            'complexportal',
            'pathwayextra',
            'kinaseextra',
        ],
    }
)

G = oc.to_annnet(
    interactions,
    source_col='source',
    target_col='target',
    edge_id_col='interaction_id',
    directed_col='is_directed',
    weight_col='curation_score',
    edge_attr_cols=['consensus_direction', 'source_database'],
)

print('shape:', G.shape)
G.views.edges().select(
    ['edge_id', 'source', 'target', 'effective_weight', 'source_database']
)

import polars as pl interactions = pl.DataFrame( { 'source': ['EGF', 'EGFR', 'EGFR', 'EGFR', 'RAS', 'MEK'], 'target': ['EGFR', 'RAS', 'RAS', 'GRB2', 'MEK', 'ERK'], 'interaction_id': [ 'EGF_EGFR', 'EGFR_RAS_primary', 'EGFR_RAS_secondary', 'EGFR_GRB2_complex', 'RAS_MEK', 'MEK_ERK', ], 'is_directed': [True, True, True, False, True, True], 'curation_score': [0.95, 0.88, 0.63, 0.76, 0.82, 0.79], 'consensus_direction': [1, 1, 1, 0, 1, 1], 'source_database': [ 'omnipath', 'omnipath', 'literature', 'complexportal', 'pathwayextra', 'kinaseextra', ], } ) G = oc.to_annnet( interactions, source_col='source', target_col='target', edge_id_col='interaction_id', directed_col='is_directed', weight_col='curation_score', edge_attr_cols=['consensus_direction', 'source_database'], ) print('shape:', G.shape) G.views.edges().select( ['edge_id', 'source', 'target', 'effective_weight', 'source_database'] )

---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
Cell In[2], line 29
      1 import polars as pl
      3 interactions = pl.DataFrame(
      4     {
      5         'source': ['EGF', 'EGFR', 'EGFR', 'EGFR', 'RAS', 'MEK'],
   (...)     26     }
     27 )
---> 29 G = oc.to_annnet(
     30     interactions,
     31     source_col='source',
     32     target_col='target',
     33     edge_id_col='interaction_id',
     34     directed_col='is_directed',
     35     weight_col='curation_score',
     36     edge_attr_cols=['consensus_direction', 'source_database'],
     37 )
     39 print('shape:', G.shape)
     40 G.views.edges().select(
     41     ['edge_id', 'source', 'target', 'effective_weight', 'source_database']
     42 )

NameError: name 'oc' is not defined

Add analysis context as slices¶

In [3]:

rows = list(G.views.edges().iter_rows(named=True))
edge_label = {
    row['edge_id']: f"{row['source']} -> {row['target']}"
    for row in rows
}
high_confidence = [
    row['edge_id']
    for row in rows
    if row['effective_weight'] >= 0.85
]
G.slices.add('high_confidence')
G.slices.add_edges('high_confidence', high_confidence)

print(
    'high-confidence interactions:',
    [edge_label[eid] for eid in sorted(G.slices.edges('high_confidence'))],
)

rows = list(G.views.edges().iter_rows(named=True)) edge_label = { row['edge_id']: f"{row['source']} -> {row['target']}" for row in rows } high_confidence = [ row['edge_id'] for row in rows if row['effective_weight'] >= 0.85 ] G.slices.add('high_confidence') G.slices.add_edges('high_confidence', high_confidence) print( 'high-confidence interactions:', [edge_label[eid] for eid in sorted(G.slices.edges('high_confidence'))], )

---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
Cell In[3], line 1
----> 1 rows = list(G.views.edges().iter_rows(named=True))
      2 edge_label = {
      3     row['edge_id']: f"{row['source']} -> {row['target']}"
      4     for row in rows
      5 }
      6 high_confidence = [
      7     row['edge_id']
      8     for row in rows
      9     if row['effective_weight'] >= 0.85
     10 ]

NameError: name 'G' is not defined

Draw the prior network¶

In [4]:

from annnet.utils import plotting

plotting.plot(
    G,
    backend='graphviz',
    show_edge_labels=True,
    edge_label_keys=['source_database'],
)

from annnet.utils import plotting plotting.plot( G, backend='graphviz', show_edge_labels=True, edge_label_keys=['source_database'], )

---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
Cell In[4], line 4
      1 from annnet.utils import plotting
      3 plotting.plot(
----> 4     G,
      5     backend='graphviz',
      6     show_edge_labels=True,
      7     edge_label_keys=['source_database'],
      8 )

NameError: name 'G' is not defined

The client hands back a graph that keeps confidence, provenance and every downstream context in one object. Where the table came from is the concern of the client, and what the graph does with it is the concern of AnnNet.