Gene Regulatory Network Tutorial¶

In this tutorial we run through GRN graph construction in Graphein. This works by providing a gene_list containing gene names. The interactions between these genes are retrieved from two possible sources:

TRRUST
RegNetwork

This is controlled by providing an edge construction function (add_trrust_edges and add_regnetwork_edges) respectively.

[1]:

import networkx as nx
from functools import partial
from typing import List
import matplotlib.pyplot as plt

import logging
logging.basicConfig(level=logging.INFO)
from easydev import Logging as elogging
elogging.level = "CRITICAL"

from graphein.grn.edges import add_regnetwork_edges, add_trrust_edges

INFO:numexpr.utils:Note: NumExpr detected 64 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
INFO:numexpr.utils:NumExpr defaulting to 8 threads.

Config Objects¶

Global parameters are stored in Config objects. We have a global GRNGraphConfig that contains a TRRUSTConfig and a RegNetworkConfig for parameters relating to each of the two sources.

[2]:

# Initialise Config Object
from graphein.grn.config import GRNGraphConfig
config = GRNGraphConfig()
print(f"GRN Config: \n{config} \n")
print(f"TRRUST Config: \n{config.trrust_config} \n")
print(f"RegNetwork Config: \n{config.regnetwork_config}")

GRN Config:
kwargs={} trrust_config=TRRUSTConfig(filtering_functions=None, root_dir=None, kwargs=None) regnetwork_config=RegNetworkConfig(filtering_functions=None, root_dir=None, kwargs=None)

TRRUST Config:
filtering_functions=None root_dir=None kwargs=None

RegNetwork Config:
filtering_functions=None root_dir=None kwargs=None

Gene List¶

This is the list of genes that we wish to construct a GRN graph for

[3]:

gene_list: List[str] = ["AATF", "MYC", "USF1", "SP1", "TP53", "DUSP1"]

Edge Annotation Functions¶

These are functions that determine labelling of edges with additional metadata. Here we define one which assigs: * "+" to upregulatory interactions * "-" to downregulatory interactions * "?" to interactons of unknown directionality

Edge Annotation functions take in two nodes (u, v) and the edge data (d)

[4]:

# This function defines the labelling of regulatory directions
def edge_ann_fn(u, v, d):
    if "+" in d["regtype"]:
        d["regtype"] = "+"
    elif "-" in d["regtype"]:
        d["regtype"] = "-"
    elif "?" in d["regtype"]:
        d["regtype"] = "?"

Graph Construction¶

[6]:

from graphein.grn.graphs import compute_grn_graph
from graphein.grn.features.node_features import add_sequence_to_nodes

g = compute_grn_graph(
    gene_list=gene_list,
    edge_construction_funcs=[
        partial(add_trrust_edges, trrust_filtering_funcs=config.trrust_config.filtering_functions),
        partial(add_regnetwork_edges, regnetwork_filtering_funcs=config.regnetwork_config.filtering_functions),
    ],
    node_annotation_funcs=[add_sequence_to_nodes],#, molecular_weight],
    edge_annotation_funcs=[edge_ann_fn],
)

[01/09/23 19:06:00] DEBUG    Added 6 nodes to graph                              graphs.py:90

WARNING [bioservices.HGNC:687]:  URL of the services contains a double //.Check your URL and remove trailing /
WARNING:bioservices.HGNC:URL of the services contains a double //.Check your URL and remove trailing /
WARNING [bioservices.HGNC:687]:  URL of the services contains a double //.Check your URL and remove trailing /
WARNING:bioservices.HGNC:URL of the services contains a double //.Check your URL and remove trailing /
WARNING [bioservices.HGNC:687]:  URL of the services contains a double //.Check your URL and remove trailing /
WARNING:bioservices.HGNC:URL of the services contains a double //.Check your URL and remove trailing /
WARNING [bioservices.HGNC:687]:  URL of the services contains a double //.Check your URL and remove trailing /
WARNING:bioservices.HGNC:URL of the services contains a double //.Check your URL and remove trailing /
WARNING [bioservices.HGNC:687]:  URL of the services contains a double //.Check your URL and remove trailing /
WARNING:bioservices.HGNC:URL of the services contains a double //.Check your URL and remove trailing /
WARNING [bioservices.HGNC:687]:  URL of the services contains a double //.Check your URL and remove trailing /
WARNING:bioservices.HGNC:URL of the services contains a double //.Check your URL and remove trailing /
WARNING [bioservices.HGNC:687]:  URL of the services contains a double //.Check your URL and remove trailing /
WARNING:bioservices.HGNC:URL of the services contains a double //.Check your URL and remove trailing /
WARNING [bioservices.HGNC:687]:  URL of the services contains a double //.Check your URL and remove trailing /
WARNING:bioservices.HGNC:URL of the services contains a double //.Check your URL and remove trailing /
WARNING [bioservices.HGNC:687]:  URL of the services contains a double //.Check your URL and remove trailing /
WARNING:bioservices.HGNC:URL of the services contains a double //.Check your URL and remove trailing /
WARNING [bioservices.HGNC:687]:  URL of the services contains a double //.Check your URL and remove trailing /
WARNING:bioservices.HGNC:URL of the services contains a double //.Check your URL and remove trailing /
WARNING [bioservices.HGNC:687]:  URL of the services contains a double //.Check your URL and remove trailing /
WARNING:bioservices.HGNC:URL of the services contains a double //.Check your URL and remove trailing /
WARNING [bioservices.HGNC:687]:  URL of the services contains a double //.Check your URL and remove trailing /
WARNING:bioservices.HGNC:URL of the services contains a double //.Check your URL and remove trailing /
/home/atj39/github/graphein/graphein/grn/parse_trrust.py:128: SettingWithCopyWarning:
A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  df["regtype"].replace(

[01/09/23 19:06:13] DEBUG    Added 10 trrust interaction edges                    edges.py:94

/home/atj39/github/graphein/graphein/grn/parse_regnetwork.py:235: SettingWithCopyWarning:
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  df["source"] = "RegNetwork"
/home/atj39/github/graphein/graphein/grn/parse_regnetwork.py:238: SettingWithCopyWarning:
A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  df["regtype"].replace(

                    DEBUG    Added 23 regnetwork interaction edges                edges.py:94

[7]:

for n,d in g.nodes(data=True):
    print(n, d)

AATF {'gene_id': 'AATF', 'uniprot_ids': ['Q9NY61'], 'sequence_Q9NY61': 'MAGPQPLALQLEQLLNPRPSEADPEADPEEATAARVIDRFDEGEDGEGDFLVVGSIRKLASASLLDTDKRYCGKTTSRKAWNEDHWEQTLPGSSDEEISDEEGSGDEDSEGLGLEEYDEDDLGAAEEQECGDHRESKKSRSHSAKTPGFSVQSISDFEKFTKGMDDLGSSEEEEDEESGMEEGDDAEDSQGESEEDRAGDRNSEDDGVVMTFSSVKVSEEVEKGRAVKNQIALWDQLLEGRIKLQKALLTTNQLPQPDVFPLFKDKGGPEFSSALKNSHKALKALLRSLVGLQEELLFQYPDTRYLVDGTKPNAGSEEISSEDDELVEEKKQQRRRVPAKRKLEMEDYPSFMAKRFADFTVYRNRTLQKWHDKTKLASGKLGKGFGAFERSILTQIDHILMDKERLLRRTQTKRSVYRVLGKPEPAAQPVPESLPGEPEILPQAPANAHLKDLDEEIFDDDDFYHQLLRELIERKTSSLDPNDQVAMGRQWLAIQKLRSKIHKKVDRKASKGRKLRFHVLSKLLSFMAPIDHTTMNDDARTELYRSLFGQLHPPDEGHGD'}
MYC {'gene_id': 'MYC', 'uniprot_ids': ['P01106'], 'sequence_P01106': 'MPLNVSFTNRNYDLDYDSVQPYFYCDEEENFYQQQQQSELQPPAPSEDIWKKFELLPTPPLSPSRRSGLCSPSYVAVTPFSLRGDNDGGGGSFSTADQLEMVTELLGGDMVNQSFICDPDDETFIKNIIIQDCMWSGFSAAAKLVSEKLASYQAARKDSGSPNPARGHSVCSTSSLYLQDLSAAASECIDPSVVFPYPLNDSSSPKSCASQDSSAFSPSSDSLLSSTESSPQGSPEPLVLHEETPPTTSSDSEEEQEDEEEIDVVSVEKRQAPGKRSESGSPSAGGHSKPPHSPLVLKRCHVSTHQHNYAAPPSTRKDYPAAKRVKLDSVRVLRQISNNRKCTSPRSSDTEENVKRRTHNVLERQRRNELKRSFFALRDQIPELENNEKAPKVVILKKATAYILSVQAEEQKLISEEDLLRKRREQLKHKLEQLRNSCA'}
USF1 {'gene_id': 'USF1', 'uniprot_ids': ['P22415'], 'sequence_P22415': 'MKGQQKTAETEEGTVQIQEGAVATGEDPTSVAIASIQSAATFPDPNVKYVFRTENGGQVMYRVIQVSEGQLDGQTEGTGAISGYPATQSMTQAVIQGAFTSDDAVDTEGTAAETHYTYFPSTAVGDGAGGTTSGSTAAVVTTQGSEALLGQATPPGTGQFFVMMSPQEVLQGGSQRSIAPRTHPYSPKSEAPRTTRDEKRRAQHNEVERRRRDKINNWIVQLSKIIPDCSMESTKSGQSKGGILSKACDYIQELRQSNHRLSEELQGLDQLQLDNDVLRQQVEDLKNKNLLLRAQLRHHGLEVVIKNDSN'}
SP1 {'gene_id': 'SP1', 'uniprot_ids': ['P08047'], 'sequence_P08047': 'MSDQDHSMDEMTAVVKIEKGVGGNNGGNGNGGGAFSQARSSSTGSSSSTGGGGQESQPSPLALLAATCSRIESPNENSNNSQGPSQSGGTGELDLTATQLSQGANGWQIISSSSGATPTSKEQSGSSTNGSNGSESSKNRTVSGGQYVVAAAPNLQNQQVLTGLPGVMPNIQYQVIPQFQTVDGQQLQFAATGAQVQQDGSGQIQIIPGANQQIITNRGSGGNIIAAMPNLLQQAVPLQGLANNVLSGQTQYVTNVPVALNGNITLLPVNSVSAATLTPSSQAVTISSSGSQESGSQPVTSGTTISSASLVSSQASSSSFFTNANSYSTTTTTSNMGIMNFTTSGSSGTNSQGQTPQRVSGLQGSDALNIQQNQTSGGSLQAGQQKEGEQNQQTQQQQILIQPQLVQGGQALQALQAAPLSGQTFTTQAISQETLQNLQLQAVPNSGPIIIRTPTVGPNGQVSWQTLQLQNLQVQNPQAQTITLAPMQGVSLGQTSSSNTTLTPIASAASIPAGTVTVNAAQLSSMPGLQTINLSALGTSGIQVHPIQGLPLAIANAPGDHGAQLGLHGAGGDGIHDDTAGGEEGENSPDAQPQAGRRTRREACTCPYCKDSEGRGSGDPGKKKQHICHIQGCGKVYGKTSHLRAHLRWHTGERPFMCTWSYCGKRFTRSDELQRHKRTHTGEKKFACPECPKRFMRSDHLSKHIKTHQNKKGGPGVALSVGTLPLDSGAGSEGSGTATPSALITTNMVAMEAICPEGIARLANSGINVMQVADLQSINISGNGF'}
TP53 {'gene_id': 'TP53', 'uniprot_ids': ['P04637'], 'sequence_P04637': 'MEEPQSDPSVEPPLSQETFSDLWKLLPENNVLSPLPSQAMDDLMLSPDDIEQWFTEDPGPDEAPRMPEAAPPVAPAPAAPTPAAPAPAPSWPLSSSVPSQKTYQGSYGFRLGFLHSGTAKSVTCTYSPALNKMFCQLAKTCPVQLWVDSTPPPGTRVRAMAIYKQSQHMTEVVRRCPHHERCSDSDGLAPPQHLIRVEGNLRVEYLDDRNTFRHSVVVPYEPPEVGSDCTTIHYNYMCNSSCMGGMNRRPILTIITLEDSSGNLLGRNSFEVRVCACPGRDRRTEEENLRKKGEPHHELPPGSTKRALPNNTSSSPQPKKKPLDGEYFTLQIRGRERFEMFRELNEALELKDAQAGKEPGGSRAHSSHLKSKKGQSTSRHKKLMFKTEGPDSD'}
DUSP1 {'gene_id': 'DUSP1', 'uniprot_ids': ['P28562'], 'sequence_P28562': 'MVMEVGTLDAGGLRALLGERAAQCLLLDCRSFFAFNAGHIAGSVNVRFSTIVRRRAKGAMGLEHIVPNAELRGRLLAGAYHAVVLLDERSAALDGAKRDGTLALAAGALCREARAAQVFFLKGGYEAFSASCPELCSKQSTPMGLSLPLSTSVPDSAESGCSSCSTPLYDQGGPVEILPFLYLGSAYHASRKDMLDALGITALINVSANCPNHFEGHYQYKSIPVEDNHKADISSWFNEAIDFIDSIKNAGGRVFVHCQAGISRSATICLAYLMRTNRVKLDEAFEFVKQRRSIISPNFSFMGQLLQFESQVLAPHCSAEAGSPAMAVLDRGTSTTTVFNFPVSIPVHSTNSALSYLQSPITTSPSC'}

Plotting the Graph¶

We now wish to visualise the graph. We can do this with the built in visualisation in NetworkX

[8]:

# Here we define a colouring scheme for the graph
edge_colors = [
    "r"
    if g[u][v]["kind"] == {"trrust"}
    else "b"
    if g[u][v]["kind"] == {"regnetwork"}
    else "y"
    for u, v in g.edges()
]

print(g)

pos = nx.spring_layout(g)
nx.draw(g, pos=pos, with_labels=True, edge_color=edge_colors)
edge_labels = {(u, v): g[u][v]["regtype"] for u, v in g.edges}
nx.draw_networkx_edge_labels(g, pos=pos, edge_labels=edge_labels)
plt.show()

DiGraph with 6 nodes and 22 edges

../_images/notebooks_grn_tutorial_12_1.png

[9]:

print(g.edges(data=True))

[('AATF', 'TP53', {'kind': {'trrust'}, 'regtype': '?'}), ('AATF', 'MYC', {'kind': {'trrust'}, 'regtype': '+'}), ('AATF', 'SP1', {'kind': {'regnetwork'}, 'regtype': '?'}), ('MYC', 'TP53', {'kind': {'trrust', 'regnetwork'}, 'regtype': '+'}), ('MYC', 'SP1', {'kind': {'regnetwork'}, 'regtype': '?'}), ('MYC', 'DUSP1', {'kind': {'regnetwork'}, 'regtype': '?'}), ('MYC', 'MYC', {'kind': {'regnetwork'}, 'regtype': '?'}), ('USF1', 'DUSP1', {'kind': {'regnetwork'}, 'regtype': '?'}), ('USF1', 'SP1', {'kind': {'regnetwork'}, 'regtype': '?'}), ('USF1', 'USF1', {'kind': {'regnetwork'}, 'regtype': '?'}), ('USF1', 'AATF', {'kind': {'regnetwork'}, 'regtype': '?'}), ('USF1', 'TP53', {'kind': {'regnetwork'}, 'regtype': '?'}), ('SP1', 'MYC', {'kind': {'trrust', 'regnetwork'}, 'regtype': '+'}), ('SP1', 'SP1', {'kind': {'regnetwork'}, 'regtype': '?'}), ('SP1', 'TP53', {'kind': {'regnetwork'}, 'regtype': '?'}), ('SP1', 'USF1', {'kind': {'regnetwork'}, 'regtype': '?'}), ('SP1', 'AATF', {'kind': {'regnetwork'}, 'regtype': '?'}), ('SP1', 'DUSP1', {'kind': {'regnetwork'}, 'regtype': '?'}), ('TP53', 'MYC', {'kind': {'trrust', 'regnetwork'}, 'regtype': '-'}), ('TP53', 'TP53', {'kind': {'trrust', 'regnetwork'}, 'regtype': '+'}), ('TP53', 'DUSP1', {'kind': {'trrust', 'regnetwork'}, 'regtype': '+'}), ('TP53', 'SP1', {'kind': {'regnetwork'}, 'regtype': '?'})]