From 10a59086a0a49b66d02f8c27ea2ad2e754549301 Mon Sep 17 00:00:00 2001
From: Fontrodona Nicolas <nicolas.fontrodona@ens-lyon.fr>
Date: Mon, 23 Nov 2020 10:05:26 +0100
Subject: [PATCH] src/find_interaction_cluster/nt_and_community.py:
modification of multiple_nt_lmm_launcher launcher
---
.../nt_and_community.py | 54 ++++++++-----------
1 file changed, 22 insertions(+), 32 deletions(-)
diff --git a/src/find_interaction_cluster/nt_and_community.py b/src/find_interaction_cluster/nt_and_community.py
index 8e7b6394..d95347d5 100644
--- a/src/find_interaction_cluster/nt_and_community.py
+++ b/src/find_interaction_cluster/nt_and_community.py
@@ -22,7 +22,6 @@ from .community_finder import get_projects
from ..logging_conf import logging_def
from itertools import product
import multiprocessing as mp
-from .sf_and_communities import get_key
def get_nt_frequency(cnx: sqlite3.Connection, list_ft: List[str],
@@ -264,6 +263,7 @@ def get_stat_nt_communities(df: pd.DataFrame, project: str, weight: int,
res['pval'] = pval
nt_ctrl_table = noutfile.parent / noutfile.name.replace("_stat.txt",
"_ctrl.txt")
+ print(df.head())
ndf = create_ctrl_community(df, nt_ctrl_table, feature, region)
sum_df = lmm_maker_summary(ndf, outfile, nt)
outfile_ctrl = ConfigGraph.get_community_file(project, weight,
@@ -334,44 +334,34 @@ def multiple_nt_lmm_launcher(ps: int,
project = get_projects(global_weight, project)
nt_list = ["A", "C", "G", "T", "S", "W"]
condition = list(product([project], [weight], nt_list))
- processes = {}
+ processes = []
pool = mp.Pool(processes=min(ps, len(condition)))
logging.debug("Calculating stats...")
- dic_df = {}
for project, weight, nt in condition:
- ckey = get_key(project, weight)
- if ckey in dic_df:
- df = dic_df[ckey]
- else:
- df = create_dataframe(project, weight, global_weight, same_gene,
- feature)
- nfile_table = ConfigGraph.get_community_file(
+ df = create_dataframe(project, weight, global_weight, same_gene,
+ feature)
+ nfile_table = ConfigGraph.get_community_file(
project, weight, global_weight, same_gene, feature,
f"_nt_table.txt", "sf_community_enrichment")
- df.to_csv(nfile_table, sep="\t", index=False)
- dic_df[ckey] = df
+ df.to_csv(nfile_table, sep="\t", index=False)
+
+
args = [df, project, weight, global_weight, same_gene, nt, feature,
region]
- if ckey in processes:
- processes[ckey].append(
- pool.apply_async(get_stat_nt_communities, args))
- else:
- processes[ckey] = [pool.apply_async(get_stat_nt_communities, args)]
- for p, value in processes.items():
- project, weight = p.split("_")
- results = [p.get(timeout=None) for p in value]
- pool.close()
- pool.join()
- fdf = pd.DataFrame(results)
- fdf["padj"] = multipletests(fdf['pval'].values, method='fdr_bh')[1]
- outfile = ConfigGraph.get_community_file(project, weight,
- global_weight,
- same_gene, feature,
- f"lmm-nt_stat.txt",
- "sf_community_enrichment")
- nfolder = outfile.parent / "nt_analysis"
- noutfile = nfolder / outfile.name
- fdf.to_csv(noutfile, sep="\t", index=False)
+ processes.append(pool.apply_async(get_stat_nt_communities, args))
+ results = [p.get(timeout=None) for p in processes]
+ pool.close()
+ pool.join()
+ fdf = pd.DataFrame(results)
+ fdf["padj"] = multipletests(fdf['pval'].values, method='fdr_bh')[1]
+ outfile = ConfigGraph.get_community_file(project, weight,
+ global_weight,
+ same_gene, feature,
+ f"lmm-nt_stat.txt",
+ "sf_community_enrichment")
+ nfolder = outfile.parent / "nt_analysis"
+ noutfile = nfolder / outfile.name
+ fdf.to_csv(noutfile, sep="\t", index=False)
if __name__ == "__main__":
--
GitLab