def enrichment_analysis(self, library, output):
'''
Saves the results of enrichment analysis
Attributes:
-----------
library - Enrichr library to be used. Recommendations:
- 'GO_Molecular_Function_2018'
- 'GO_Biological_Process_2018'
- 'GO_Cellular_Component_2018'
for more options check available libraries by typing gseapy.get_library_name()
output - directory name where results should be saved
'''
libs = gseapy.get_library_name()
assert library in libs, "the library is not available, check gseapy.get_library_name() for available options"
assert (self.convert == True) or (
self.origID == "symbol"
), "EnrichR accepts only gene names as an input, thus please set 'convert' to True and indicate the original gene ID"
genes1_name = [self.mapping[x] for x in self.genes1]
genes2_name = [self.mapping[x] for x in self.genes2]
all_genes_names = genes1_name + genes2_name
gseapy.enrichr(gene_list=all_genes_names,
description='pathway',
gene_sets=library,
cutoff=0.05,
outdir=output)
def gsea(genes,
description='',
out='./',
sleeptime=1,
sleep=False,
gsets=[
'GO_Biological_Process_2018', 'KEGG_2019_Human',
'WikiPathways_2019_Human'
]):
"""
genes (list of str): gene symbols
description (str): name for enrichment report
sleeptime (int): length of wait time between each query
(overloading server causes connection to be cut)
"""
if sleep:
for gset in gsets:
time.sleep(sleeptime)
gseapy.enrichr(gene_list=genes,
description=description,
gene_sets=gset,
outdir=out)
else:
gseapy.enrichr(gene_list=genes,
description=description,
gene_sets=gsets,
outdir=out)
return
def enrichr_validation(gene_list,
gene_rank=None,
outdir="validation_results",
gene_sets='KEGG_2016'):
"""
Perform the enrichr tool (http://amp.pharm.mssm.edu/Enrichr/)
Enrichment of a gene list
Args:
-gene_list (list): Gene list to analyze
-gene_rank (list): Ranking of the genes (according to a scoring function)
-outdir (str): Location to save the files
-gene_sets (str): Gene set to use for the enrichment
"""
if gene_rank == None:
enr = gp.enrichr(gene_list=gene_list,
description='pathway',
gene_sets='KEGG_2016',
outdir='test',
cutoff=0.05,
format='png')
else:
assert type(gene_rank) == list, "please provide gene_rank as a list"
rnk = pd.DataFrame(np.array([gene_list, gene_rank]).T,
columns=['gene', 'score'])
enr = gp.enrichr(gene_list=rnk,
description='pathway',
gene_sets='KEGG_2016',
outdir=outdir,
cutoff=0.05,
format='png')
result = enr.res2d[enr.res2d["Adjusted P-value"] < pvalue]
return result
def res(x_one, y_one, top):
array_one = []
enr_x_one = None
try:
enr_x_one = gp.enrichr(gene_list=x_one,
gene_sets=lib,
organism='Human',
cutoff=0.05).results[[
'Term', 'P-value'
]].head(top).values.tolist()
except Exception:
pass
if enr_x_one is not None and len(enr_x_one) > 0:
enr_y_one = None
try:
enr_y_one = gp.enrichr(gene_list=y_one,
gene_sets=lib,
organism='Human',
cutoff=1.0).results[['Term', 'P-value']]
except Exception:
pass
if enr_y_one is not None:
for term in enr_x_one:
pair = enr_y_one.loc[enr_y_one['Term'] == term[0]]
if pair is not None and pair.shape[0] > 0:
pair = pair.iloc[0].values.tolist()
array_one.append((term[1], pair[1]))
return array_one
def enrichr(gene_list,
description,
out_dir,
scan=None,
max_terms=10,
figsize=(12, 6),
run_main=False):
'''
Performs GO Molecular Function, GO Biological Process and KEGG enrichment on a gene list.
Uses enrichr.
Inputs
------
gene_list: list of genes to perform enrichment on
description: string description for title
out_dir: output director
scan: dictionary with additional enrichr dbs to scan (http://amp.pharm.mssm.edu/Enrichr/#stats)
max_terms: limit return plot to this max
load: load results
figsize: change fig size
Returns
-------
None
'''
out_dir = make_folder(out_dir)
testscan = {
'KEGG': 'KEGG_2016',
'GO_biological_process': 'GO_Biological_Process_2017b',
'ChIP-X_Consensus_TFs': 'ENCODE_and_ChEA_Consensus_TFs_from_ChIP-X',
'ChEA': 'ChEA_2016',
'OMIM_Disease': 'OMIM_Disease'
}
if isinstance(scan, dict):
testscan = {**testscan, **scan}
for nick, name in testscan.items():
gseapy.enrichr(gene_list=gene_list,
figsize=figsize,
top_term=max_terms,
description=f'{description}_{nick}',
gene_sets=name,
outdir=out_dir,
format='png')
out_result(f'{out_dir}{nick}.{name}.enrichr.reports.png',
f'Enrichr: {nick} for {description}',
run_main=run_main)
out_list = pd.DataFrame({'Gene Name': gene_list},
index=range(len(gene_list)))
out_list.to_excel(f'{out_dir}{description}_genes.xlsx', index=None)
def adj_enrich_score(ranks, gene_set_file, gene_sets, cutoff=.5):
clusters = ranks.columns.tolist()
# Just to get pathway names
nes_df = pd.DataFrame(index=gene_sets)
for cluster in clusters:
print("computing for cluster %s" % cluster)
genes = ranks.index[(ranks[cluster] > cutoff).tolist()]
#print(genes)
print(len(genes))
enr = gp.enrichr(
gene_list=genes.tolist(),
# or gene_list=glist
description='test_name',
gene_sets=gene_set_file,
outdir='../test/enrichr_kegg',
cutoff=1 # test dataset, use lower value of range(0,1)
)
try:
enr.results.index = enr.results["Term"]
#print(enr.results.head())
#print(nes_df.head())
nes_df[cluster] = enr.results["Adjusted P-value"]
except KeyError:
print(enr.results.columns)
#print(nes_df.head())
#print("number of positive nes scores %d" % (nes_df[cluster] > 0).sum())
return nes_df
def save_enrichment(x):
lib = gp.get_library_name('Human')
with open('gensets.txt', 'w') as f:
for item in range(len(lib)):
f.write("%s %s\n" % (item, lib[item]))
# lib = lib[49: 54]
lib = lib[53]
files = [(1, x+"/gcn-hom-hom.csv"), (2, x+"/gcn-hom-onto.csv"),
(3, x+"/gcn-onto-onto.csv"), (4, x+"/gae-hom-hom.csv"),
(5, x+"/gae-hom-onto.csv"), (6, x+"/gae-onto-onto.csv")]
df = pd.DataFrame()
writer = pd.ExcelWriter('enrich-cluster/full-results.xlsx')
for key, file in files:
print(file)
cluster_data = read_file_2(file)
for i in cluster_data:
try:
enr = gp.enrichr(gene_list=list(cluster_data[i][2]), gene_sets=lib, organism='Human', cutoff=0.05).results
except:
pass
enr['model'] = key
enr['cluster'] = i
df = df.append(enr)
df = df[(df['P-value'] < 0.05)]
df.to_excel(writer, sheet_name="sheet1")
writer.save()
def enrichr_go_bp(symbols=None, cutoff=0.05):
dummy_directory = 'biopipe-enrichr-dummy'
try:
enrichr_result = gseapy.enrichr(
gene_list=symbols,
gene_sets='GO_Biological_Process_2017b',
outdir=dummy_directory,
no_plot=True)
except Exception:
print(
'An error occurred during running enrichr. Please try again later.'
)
sys.exit(1)
finally:
shutil.rmtree(dummy_directory)
result_dataframe = enrichr_result.res2d
p_value_filtered_result = result_dataframe[
result_dataframe['Adjusted P-value'] < cutoff]
transformed_p_values = [
-np.log10(p)
for p in p_value_filtered_result['Adjusted P-value'].values
]
terms = p_value_filtered_result['Term'].values
data = list(zip(terms, transformed_p_values))[:10]
terminal_bar_chart(data,
title='%s [-log10(p)]' % enrichr_result.gene_sets,
sort=True)
def __init__(self, data:H5COUNTS, path="data/interim/",
threshold=0.05,
gene_sets=['GO_Biological_Process_2018', 'GO_Cellular_Component_2018', 'GO_Molecular_Function_2018'],
tumor_ids=[1, 2, 3, 4, 5, 6, 7, 8]):
self.gsea_table = pd.DataFrame()
self.data = data
for tumor_id in tumor_ids:
de_genes_tumor_df = pd.read_csv(path+"MK_genes_TUMOR{}.csv".format(tumor_id))
de_genes_by_cluster = de_genes_tumor_df.groupby("cluster")["gene"].apply(lambda x: "|".join(x.unique()))
tumor_name = data.id2tumor[tumor_id]
print("Running GSEA for tumor", tumor_name)
for cluster in de_genes_by_cluster.index:
DE_gene_list = de_genes_by_cluster[cluster].split("|")
tumor_cluster = tumor_name + "_" + str(cluster)
enr = gp.enrichr(gene_list=DE_gene_list,
gene_sets=gene_sets,
no_plot=True,
cutoff=0.05 # test dataset, use lower value from range(0,1)
)
if threshold:
enr.results = enr.results[enr.results["Adjusted P-value"] < threshold]
enr_results = enr.results.set_index("Term")
for geneset in enr_results.index:
self.gsea_table.loc[geneset, tumor_cluster] = enr_results.loc[geneset, "Adjusted P-value"]
self.gsea_table = self.gsea_table.T
self.gsea_table.index = pd.MultiIndex.from_tuples(self.gsea_table.index.str.split("_", expand=True),
names=["tumor", "cluster"])
def _enrichr(self, category, background=None, verbose=True):
if background is None:
background = self.background
if isinstance(category, list):
gene_list = category
else:
assert category in ['up', 'down', 'all']
gene_list = list(self.rnadiff.gene_lists[category])
if self.mapper is not None:
logger.info("Input gene list of {} ids".format(len(gene_list)))
#gene_list = [x.replace("gene:", "") for x in gene_list]
identifiers = self.mapper.loc[gene_list]['name'].drop_duplicates(
).values
logger.info("Mapped gene list of {} ids".format(len(identifiers)))
gene_list = list(identifiers)
enr = gseapy.enrichr(gene_list=gene_list,
gene_sets=self.gene_sets,
verbose=verbose,
background=background,
outdir="test",
no_plot=True)
return enr
def save_enrichment_set():
lib = gp.get_library_name('Human')
lib = lib[53]
files = [("gcn-hom-hom", "enrich/gcn-hom-hom.csv"),
("gcn-hom-onto", "enrich/gcn-hom-onto.csv"),
("gcn-onto-onto", "enrich/gcn-onto-onto.csv"),
("gae-hom-hom", "enrich/gae-hom-hom.csv"),
("gae-hom-onto", "enrich/gae-hom-onto.csv"),
("gae-onto-onto", "enrich/gae-onto-onto.csv")]
enrich_set = {}
for key, file in files:
print(file)
cluster_data = read_file(file)
for i in cluster_data:
print(len(cluster_data[i][2]))
try:
enr = gp.enrichr(gene_list=list(cluster_data[i][2])[:1000], gene_sets=lib, organism='Human', cutoff=0.05).results
name = key + "-" + str(i)
term = enr['Term'].to_list()
enrich_set[name] = term
# print(i)
print(enr)
except:
pass
write_file("enrich-cluster/full_result_dic.csv", enrich_set)
def vec_enrich(vec, gene_ids, quantile, gene_sets):
ind = np.quantile(vec, quantile) > vec
print("... {} features selected...".format(sum(ind)))
genes = gene_ids[ind]
# remove ens id version
genes = [re.sub("\\..*$", "", g) for g in genes]
print("Mapping to gene names...")
# map ens ids to gene symbols
bm = Biomart()
bm_result = bm.query(
dataset="hsapiens_gene_ensembl",
attributes=[
"ensembl_gene_id",
"external_gene_name",
"entrezgene_id",
"go_id",
],
filters={"ensembl_gene_id": genes},
)
gene_symbols = list(bm_result["external_gene_name"].unique())
print("Calculating enrichment...")
enr = gp.enrichr(
gene_list=gene_symbols,
gene_sets=gene_sets,
organism="Human",
cutoff=0.05,
)
return enr
def enrichr_test(direction="pos", significance=True):
genes = np.array(rnk["gene"])
pvals = np.array(rnk["pval"])
fcs = np.array(rnk["fc"])
if direction == "pos":
if significance:
hits = set(genes[np.logical_and(fcs > 0, pvals < alpha)])
else:
hits = set(genes[:top])
else:
if significance:
hits = set(genes[np.logical_and(fcs < 0, pvals < alpha)])
else:
hits = set(genes[-top:])
if significance:
sur = "sign05"
else:
sur = "top100"
outpath = OUTPATH + "/enrichr_%s_%s/" % (direction, sur)
if not os.path.exists(outpath): os.mkdir(outpath)
enr = gseapy.enrichr(
gene_list=list(hits),
description="%s_%s_%s" % (filename, direction, sur),
gene_sets=gene_sets,
outdir=outpath,
cutoff=0.5 # only used for plotting.
)
def get_ontology_df(
topic,
cutoff=0.05,
threshhold=5e-1,
gene_sets=[
'GO_Molecular_Function_2018', 'GO_Biological_Process_2018',
'GO_Cellular_Component_2018', 'Human_Phenotype_Ontology',
'GTEx_Tissue_Sample_Gene_Expression_Profiles_up',
'GTEx_Tissue_Sample_Gene_Expression_Profiles_down',
'Tissue_Protein_Expression_from_Human_Proteome_Map',
'KEGG_2019_Human', 'NCI-60_Cancer_Cell_Lines'
],
background=None):
"""
:param topic: list of genes
:param background: enrichment test background
:param cutoff: Enrichments cutoff
:param threshhold: threshold on Adjusted P-value
:return:
"""
sets = ','.join(gene_sets)
if background is None:
background = 'hsapiens_gene_ensembl'
topic = [g for g in topic if str(g) != 'nan']
gene_ontology = gs.enrichr(list(topic),
gene_sets=sets,
cutoff=cutoff,
background=background).results
return gene_ontology[gene_ontology['Adjusted P-value'] < threshhold][[
'Term', 'Adjusted P-value', 'Gene_set'
]]
def perform_GO_enrichment_analysis(inputGenes, geneDescription, threshold):
targetGeneSet = "GO_Biological_Process_2015" #"KEGG_2016","Reactome_2013","GO_Molecular_Function_2015","WikiPathways_2013"
if os.path.exists('./gene_set_enrichment_analysis') == False:
os.makedirs('./gene_set_enrichment_analysis')
enr = gp.enrichr(gene_list=inputGenes,
description=geneDescription,
gene_sets=targetGeneSet,
outdir='./gene_set_enrichment_analysis',
cutoff=threshold)
enr.res2d.head()
def corr_rep_gene(adata, rep, prefix='', dims=[0,1], offset=1, layer=None, thresh=0.5, out='./', \
gsets=['GO_Biological_Process_2018', 'KEGG_2019_Human', 'WikiPathways_2019_Human']):
"""
Use gseapy.get_library_name() to see more gene set option
"""
for i in dims:
df = corr_comp_gene(adata, rep, i, offset=offset, layer=layer)
df.to_csv('{}{}_{}{}_corr_gene.csv'.format(out, rep, prefix, i + 1))
df.sort_values('R', ascending=False).to_csv(
'{}{}_{}{}_corr_gene_sorted.csv'.format(out, rep, prefix, i + 1))
df_pos = df.loc[df['R'] > thresh, :]
df_neg = df.loc[df['R'] < -thresh, :]
for sign, df in zip(['pos', 'neg'], [df_pos, df_neg]):
if len(df) > 0:
gseapy.enrichr(gene_list=list(df.index),
gene_sets=gsets,
outdir='{}{}{}_{}'.format(
out, prefix, i + 1, sign))
return
def calcu_gsea(gene_list, gmt, bg):
gene_list = list(gene_list)
enr2 = gp.enrichr(
gene_list=gene_list,
# or gene_list=glist
description='test_name',
gene_sets=gmt.term_set,
background=bg, # or the number of genes, e.g 20000
outdir=None,
cutoff=0.5, # only used for testing.
verbose=True)
return enr2.results
def gsea(homepath):
'''
Parameters
----------
`homepath` (str):
Path where you want to save all the generated files
and folders.
Return:
-------
None
Outputs:
--------
Generate a directory names enrichr
within home directory and two plot
of gene enrichement analysis using
the selected genes from panclassif
'''
warnings.filterwarnings("ignore")
# Directory
directory = "enrichr"
# Parent Directory path
parent_dir = homepath
# Path
path = os.path.join(parent_dir, directory)
if not os.path.exists(path):
os.mkdir(path)
gene = pd.read_csv(homepath + "/std_npy/unique_genes_with_frequency.csv",
header=None)
gl = []
for g in range(len(gene)):
gl.append(gene[0][g])
enr = gs.enrichr(gene_list=gl,
description='Disease',
gene_sets='DisGeNET',
outdir=homepath + '/enrichr')
# simple plotting function
from gseapy.plot import barplot, dotplot
# to save your figure, make sure that ``ofname`` is not None
barplot(enr.res2d,
title='DisGeNET',
cutoff=0.2,
ofname=homepath + '/enrichr/DisGeNET_barplot.png')
dotplot(enr.res2d,
title='DisGeNET',
cmap='viridis_r',
cutoff=0.2,
ofname=homepath + '/enrichr/DisGeNET_dotplot.png')
def enrichment_GO(gene_list,
go_mode='Bio',
organism='Human',
description='test_name',
outdir='enrichment_go',
cutoff=0.5):
'''
Gene enrichment analysis of GO
Parameters
----------
gene_list:list
The gene set to be enrichment analyzed
go_mode:str
The module of GO include:'Bio','Cell','Mole'
organism:str
Select from (human, mouse, yeast, fly, fish, worm)
description:str
The title of enrichment
outdir:str
The savedir of enrichment
cutoff:float
Show enriched terms which Adjusted P-value < cutoff.
Returns
----------
result:pandas.DataFrame
stores your last query
'''
if (go_mode == 'Bio'):
geneset = 'GO_Biological_Process_2018'
if (go_mode == 'Cell'):
geneset = 'GO_Cellular_Component_2018'
if (go_mode == 'Mole'):
geneset = 'GO_Molecular_Function_2018'
enr = gp.enrichr(
gene_list=gene_list,
gene_sets=geneset,
organism=
organism, # don't forget to set organism to the one you desired! e.g. Yeast
description=description,
outdir=outdir,
# no_plot=True,
cutoff=cutoff # test dataset, use lower value from range(0,1)
)
subp = dotplot(enr.res2d, title=description, cmap='seismic')
print(subp)
return enr.res2d
def main():
usage = "" # TODO
parser = OptionParser(usage=usage)
parser.add_option("-o", "--out_file", help="OUTPUT file")
(options, args) = parser.parse_args()
de_genes_f = args[0]
#expr_fs = args[2].split(',')
out_f = options.out_file
# Get all the DE genes
de_genes_df = pd.read_csv(de_genes_f, sep='\t', index_col=0)
de_genes = de_genes_df.index
# Perform gene set enrichment
db_to_gene_sets = {}
for db_name, db in GENE_SETS.items():
enr = gp.enrichr(
gene_list=[x.strip() for x in de_genes],
#gene_list=[x.strip() for x in de_genes]
gene_sets=[db],
background=19463,
no_plot=True,
cutoff=0.05 # test dataset, use lower value from range(0,1)
)
enr.results = enr.results[
enr.results["Adjusted P-value"] < GSEA_THRESH]
sig_terms = {
str(row[0]): float(row[1])
for row_i, row in enr.results[['Term', 'Adjusted P-value'
]].iterrows()
}
db_to_gene_sets[db_name] = sig_terms
# Create final dataframe
da = []
for db, gene_set_to_pval in db_to_gene_sets.items():
for gene_set, pval in gene_set_to_pval.items():
da.append((db, gene_set, pval))
df = pd.DataFrame(data=da,
columns=['collection', 'gene_set', 'adjusted_p_value'])
df = df.sort_values(by='adjusted_p_value', axis=0)
print('{} total enriched gene sets.'.format(len(df)))
# Write output
print('Writing to {}.'.format(out_f))
df.to_csv(out_f, index=False, sep='\t')
print('done')
def enrichr(self, gene_dict, gene_set, key, reg):
"""Perform enrichr analysis on a gene dictionary of sample group : enriched gene list, a GO term library
of gene set, and a reg varible that is either 'upreg' or 'downreg'"""
# check if gene list is empty
if not gene_dict[key]:
pass
else:
# run enrichr - if there are no genes enriched with the cutoff level, it will not generate an output
enr = gp.enrichr(gene_list=gene_dict[key],
gene_sets=[gene_set],
organism='Human',
description=key + "_" + reg,
outdir=self.input_dir + 'enrichr/' + key + "/",
cutoff=0.1)
enr_df = enr.results.copy()
return enr_df
def run_enrichr(gene_list, gene_sets):
'''
gene_list: List
List containing genes names used for the analysis
gene_sets: List
List of enrichr gene libaries to use
enr: Enrichr object
Analysis output, use "enr.results" to print table of results
'''
enr = gp.enrichr(gene_list=gene_list,
description='pathway',
gene_sets=gene_sets,
organism='Human',
cutoff=0.5)
return enr
def enrichment_KEGG(gene_list,
gene_sets=['KEGG_2019_Human'],
organism='Human',
description='test_name',
outdir='enrichment_kegg',
cutoff=0.5):
'''
Gene enrichment analysis of KEGG
Parameters
----------
gene_list:list
The gene set to be enrichment analyzed
gene_sets:list
The gene_set of enrichr library
Input Enrichr Libraries (https://maayanlab.cloud/Enrichr/#stats)
organism:str
Select from (human, mouse, yeast, fly, fish, worm)
description:str
The title of enrichment
outdir:str
The savedir of enrichment
cutoff:float
Show enriched terms which Adjusted P-value < cutoff.
Returns
----------
res:pandas.DataFrame
stores your last query
'''
enr = gp.enrichr(
gene_list=gene_list,
gene_sets=gene_sets,
organism=
organism, # don't forget to set organism to the one you desired! e.g. Yeast
description=description,
outdir=outdir,
# no_plot=True,
cutoff=cutoff # test dataset, use lower value from range(0,1)
)
subp = dotplot(enr.res2d, title=description, cmap='seismic')
print(subp)
return enr.res2d
def save_top4_csvs():
for i in files:
file_name = "enrich/" + i + ".csv"
file = read_file(file_name)
temp = {}
for j in file:
gene_list = list(file[j][2])
enr_x_one = None
try:
enr_x_one = gp.enrichr(
gene_list=gene_list,
gene_sets=lib,
organism='Human',
cutoff=0.05).results.head(10)['Term'].tolist()
temp[j] = set(enr_x_one)
except:
pass
write_file("top_4_bio_process/" + i + ".csv", temp)
def enrich_tissue(tissue_name, last_num):
communities = pickle.load(
open("results/louvain_modules_" + tissue_name + ".pkl", "rb"))
corr_mat = pd.read_pickle("data/corr_" + tissue_name + ".pkl")
community_id = 1
for community in np.unique(communities[0]):
common = np.array(corr_mat.columns)[communities[0] == community]
if len(common) <= 3:
continue
# print(community_id)
print("For community", community_id,
"(len: " + str(len(common)) + ")...")
if community_id < last_num:
community_id += 1
continue
enr = gp.enrichr(gene_list=list(common.astype('<U3')),
organism='human',
description=tissue_name + "_" + str(community_id),
gene_sets='Reactome_2016',
cutoff=0.05,
outdir='results/EnrichClass')
if enr.results.shape[0] > 0:
enr.results = enr.results[enr.results['Adjusted P-value'] < 0.05]
if enr.results.shape[0] > 0:
enr.results.to_csv("results/enrichment/" + tissue_name + "_" +
str(community_id) + "_" + str(len(common)) +
".csv")
print("Enriched!")
community_id += 1
sleep(
50
) # just to go easy on the Enrich API... (constantly getting errors after a while)
def _get_top_enrichr_term(gene_sets, libraries=[
# 'GO_Biological_Process_2018',
# 'GO_Cellular_Component_2018',
# 'GO_Molecular_Function_2018',
'KEGG_2019_Human', ],
cutoff=0.01, top_k=1):
results = []
for gene_set in gene_sets:
try:
enr = gp.enrichr(gene_list=gene_set,
gene_sets=libraries,
cutoff=cutoff,
no_plot=True, verbose=False,
)
if enr.results.shape[0] > 0:
results.append(enr.results.sort_values(by="Adjusted P-value").head_node_type(top_k))
except Exception:
pass
results = [row for row in results if row is not None]
if len(results) > 0:
return pd.concat(results)
else:
return None
import gseapy as gp import pandas as pd import RNA_expression_processing as rn import matplotlib.pyplot as plt path = 'tracks/MARGE/relativeRP/' dt = pd.read_csv(path + 'DN_RegNetwork_TLX3.csv') gl = list(dt['gene_name']) gs = 'RNA-Seq_Disease_Gene_and_Drug_Signatures_from_GEO' rs = gp.enrichr(gene_list=gl, gene_sets=gs) rn.barplot(rs.res2d, ttl=gs) plt.show()
tss = set(tlx_tss_3kb_gn)
enh = set(tlx_chrhmm_gn)
#~ from matplotlib_venn import venn2
#~ venn2([tss,enh], set_labels = ('Tlx3 in Enhancer', 'Tlx3 in Promoter'))
tss_and_enh = tss & enh
tss_or_enh = tss | enh
tss_notin_enh = tss - enh
enh_notin_tss = enh - tss
diff_enh_tss = enh ^ tss
# === Run GSEApy Enrichr
gp.enrichr(gene_list=list(tss_and_enh),
description='tss_and_enh',
gene_sets='GO_Biological_Process_2017b',
outdir='gene_lists/tlx_enh_tss3k/' + 'tss_and_enh_GO_BP_2017b')
gp.enrichr(gene_list=list(tss_or_enh),
description='tss_or_enh',
gene_sets='GO_Biological_Process_2017b',
outdir='gene_lists/tlx_enh_tss3k/' + 'tss_or_enh_GO_BP_2017b')
gp.enrichr(gene_list=list(tss_notin_enh),
description='tss_notin_enh',
gene_sets='GO_Biological_Process_2017b',
outdir='gene_lists/tlx_enh_tss3k/' + 'tss_notin_enh_GO_BP_2017b')
gp.enrichr(gene_list=list(enh_notin_tss),
description='enh_notin_tss',
gene_sets='GO_Biological_Process_2017b',
for x in pd.DataFrame.from_records(r).columns:
print("group :"+x, end = "\r")
# rank gene by importance for clusters
glist = pd.DataFrame.from_records(r)[x].tolist()
bm = Biomart()
if not os.path.exists("test"):
os.makedirs("test")
results = bm.query(dataset='hsapiens_gene_ensembl',
attributes=['external_gene_name', 'go_id'],
filters={'hgnc_symbol': glist},
# save output file
filename="test/query_"+x+".results.txt")
enr = gp.enrichr(gene_list=glist,
description='test_name',
gene_sets=['KEGG_2016'],
outdir="test/enrichr_kegg_group"+x,
cutoff=0.5 # test dataset, use lower value from range(0,1)
)
# to save your figure, make sure that ``ofname`` is not None
#barplot(enr.res2d,title='',ofname="test/enrichr_kegg_group"+x+"/bar_plot.pdf")
dotplot(enr.res2d, title='',ofname="test/enrichr_kegg_group"+x+"/dot_plot.pdf")
#pd.DataFrame.from_records(adata.uns['rank_genes_groups']['scores'])[x]
rnk= pd.concat([pd.DataFrame.from_records(r)[x],pd.DataFrame.from_records(adata.uns['rank_genes_groups']['scores'])[x]],axis=1)
rnk.columns=[0,1]
pre_res = gp.prerank(rnk=rnk, gene_sets='KEGG_2016',
processes=4,
def main():
usage = "" # TODO
parser = OptionParser(usage=usage)
parser.add_option("-t", "--de_table_f", help="DE table file")
parser.add_option("-c", "--de_table_col", help="DE table column to filter")
parser.add_option("-o", "--out_file", help="Output file")
parser.add_option("-k", "--out_kept", help="Output kept genes file")
parser.add_option("-v", "--out_venn_f", help="Venn diagram file")
(options, args) = parser.parse_args()
de_genes_f = args[0]
ec_f = args[1]
out_f = options.out_file
out_kept_f = options.out_kept
out_venn_f = options.out_venn_f
# Get all the DE genes
with open(de_genes_f, 'r') as f:
de_genes = [l.strip() for l in f]
print('{} total DE genes'.format(len(de_genes)))
# Map each gene to its fold change and filter
# by fold change
if FILTER_BY_FOLD:
ec_df = pd.read_csv(ec_f, sep='\t', index_col=0)
gene_to_fc = {gene: ec_df.loc[gene]['FC'] for gene in ec_df.index}
filtered_genes = [
gene for gene in gene_to_fc
if gene_to_fc[gene] > FC_THRESH or gene_to_fc[gene] < (1 /
FC_THRESH)
]
print('{}/{} remain after filtering by fold-change.'.format(
len(filtered_genes), len(gene_to_fc)))
if len(filtered_genes) == 0:
with open(out_f, 'w') as f:
f.write('collection\tgene_set\tadjusted_p_value')
return
# Filter batch effect if specified in options
if options.de_table_f:
de_table_f = options.de_table_f
de_table_col = options.de_table_col
de_table_df = pd.read_csv(de_table_f, sep='\t', index_col=0)
batch_effect_genes = set(de_table_df.loc[de_table_df[de_table_col] ==
1][de_table_col].index)
# Output the removed genes
present_batch_effect_genes = batch_effect_genes & set(filtered_genes)
print('Removed {} genes: {}'.format(len(present_batch_effect_genes),
present_batch_effect_genes))
filtered_genes = sorted(set(filtered_genes) - batch_effect_genes)
with open(out_kept_f, 'w') as f:
f.write('\n'.join(sorted(filtered_genes)))
print('{}/{} remain after filtering by batch-effect:'.format(
len(filtered_genes), len(gene_to_fc)))
print(filtered_genes)
if len(filtered_genes) == 0:
with open(out_f, 'w') as f:
f.write('collection\tgene_set\tadjusted_p_value')
return
# Draw Venn diagram
if 'Up' in de_genes_f:
title = r'DE genes $\bf{higher}$ in COVID-19 ICU patients'
else:
title = r'DE genes $\bf{lower}$ in COVID-19 ICU patients'
fig, ax = plt.subplots(1, 1, figsize=(4.5, 3.5))
venn2(subsets=(
len(filtered_genes),
len(batch_effect_genes),
len(present_batch_effect_genes),
),
set_labels=('COVID-19 ICU\nvs. sepsis ARDS',
'non-ICU\nvs. sepsis non-ARDS'),
set_colors=('b', 'y'),
alpha=0.5,
ax=ax)
plt.gca().set_title(title, fontsize=14)
plt.tight_layout()
plt.savefig(options.out_venn_f, format='pdf')
# Perform gene set enrichment
db_to_gene_sets = {}
for db_name, db in GENE_SETS.items():
enr = gp.enrichr(
gene_list=[x.strip() for x in filtered_genes],
gene_sets=[db],
background=19463,
no_plot=True,
cutoff=0.05 # test dataset, use lower value from range(0,1)
)
enr.results = enr.results[
enr.results["Adjusted P-value"] < GSEA_THRESH]
sig_terms = {
str(row[0]): float(row[1])
for row_i, row in enr.results[['Term', 'Adjusted P-value'
]].iterrows()
}
db_to_gene_sets[db_name] = sig_terms
# Create final dataframe
da = []
for db, gene_set_to_pval in db_to_gene_sets.items():
for gene_set, pval in gene_set_to_pval.items():
da.append((db, gene_set, pval))
df = pd.DataFrame(data=da,
columns=['collection', 'gene_set', 'adjusted_p_value'])
df = df.sort_values(by='adjusted_p_value', axis=0)
print('{} total enriched gene sets.'.format(len(df)))
# Write output
print('Writing to {}.'.format(out_f))
df.to_csv(out_f, index=False, sep='\t')
print('done')
