grnsight 5.1.0 → 6.0.4

package/database/network-database/README.md

@@ -0,0 +1,88 @@
+# Network Database
+
+All files pertaining the network database live within this directory.
+
+## The basics
+
+### Schema
+
+All network data is stored within the spring2022_network schema on our Postgres database.
+
+The schema is located within this directory at the top level in the file `schema.sql`. It defines the tables located within the spring2022_network schema. 
+
+Usage:
+To load to local database
+```
+psql postgresql://localhost/postgres -f schema.sql
+```
+To load to production database
+```
+psql <address to database> -f schema.sql
+```
+
+### Scripts
+
+All scripts live within the subdirectory `scripts`, located in the top-level of the network database directory. 
+
+Any source files required to run the scripts live within the subdirectory `source-files`, located in the top-level of the network database directory. As source files may be large, you must create this directory yourself and add any source files you need to use there. 
+
+All generated results of the scripts live in the subdirectory `script-results`, located in the top-level of the network database directory. Currently, all scripts that generate code create the directory if it does not currently exist. When adding a new script that generates resulting code, best practice is to create the script-results directory and any subdirectories if it does not exist, in order to prevent errors and snafus for recently cloned repositories.
+
+Within the scripts directory, there are the following files:
+
+- `generate_network.py`
+- `loader.py`
+- `filter_genes.py`
+- `generate_sgd_network_from_yeastract_network.py`
+
+#### Network Generator (and data preprocessor)
+
+This script (`generate_network.py`) is a two-for-one. It first uses the yeastmine service from the SGD database to query for all regulator genes relating to Saccharomyces cerevisiae. From there it gets all all of the targets for each regulator gene. We then construct two networks from these connections (a regulator by regulator matrix as well as a regulator by target matrix). We also construct the processed loader files, so that they are ready to load using `loader.py`.
+
+The resulting network matrices are located in `script-results/networks` and the resulting processed loader files are located within `script-results/processed-loader-files`
+
+Make sure to have all dependencies installed beforehand or you will recieve errors. (pip3 install intermine, tzlocal, etc. [see file for all imports]
+
+Usage: 
+```
+python3 generate_network.py
+```
+#### Database Loader
+
+This script (`loader.py`) is to be used to load your preprocessed genes into the database. 
+
+This program generates direct SQL statements from the source files generated by the network generator in order to populate a relational database with those files’ data
+
+Usage: 
+To load to local database
+```
+python3 loader.py | psql postgresql://localhost/postgres
+```
+To load to production database
+```
+python3 loader.py | psql <address to database>
+```
+
+#### Filter Genes (beta functionality, not tested)
+
+This script (`filter_genes.py`) is to be used when updating a pre-existing database. It requires you to generate a new network from yeastmine using the script.`generate_network.py`. Once you generate the network, the script will access the database get all of the genes stored within. From there it will generate a csv file of all genes in the new network that are missing from your database, and all genes that have updated their display name (standard like name). You should change the database host to your localhost if you are running your own instance of postgresql and not the production database. Once finished, you can load the updated genes list using `loader.py`. *Note:* You will have to change the `GENE_SOURCE` to be the output file of the missing genes. 
+
+**Never save the password to your database in filter_genes.py! If you want, you can set up an environment variable where you store sensitive information, but for convience you will have to enter the password yourself.**
+
+Usage: 
+```
+python3 filter_genes.py
+```
+
+
+#### Generate an SGD network from a Yeastract network
+
+This script takes a network (assumed to have data from Yeastract, but it can be any given network) and gives you a network with data queried from Yeastmine (SGD). It takes the regulators and targets from a given network file, then queries Yeastmine in order to get the regulatory connections between the genes. From there, it creates a new network using the data obtained from Yeastmine. 
+
+To use, create a folder called `source-files` within the `network-database` folder. Add your network(s) to the newly created directory. Then go into the script and change the `YEASTRACT_NETWORK` to be the path of the network you would like to convert. Run the script and your SGD network (using the same genes) will be output in the `/script-results/yeastract-to-sgd-networks` directory. Change the name of the output files by editing the `SGD_MATRIX_EXCEL` and `SGD_MATRIX` variables
+
+Usage: 
+```
+python3 generate_sgd_network_from_yeastract_network.py
+```
+

package/database/network-database/schema.sql

@@ -0,0 +1,25 @@
+CREATE TABLE spring2022_network.source (
+  time_stamp TIMESTAMP WITH TIME ZONE,
+  source VARCHAR,
+  source_display_name VARCHAR,
+  PRIMARY KEY(time_stamp, source)
+);
+
+CREATE TABLE spring2022_network.gene (
+  gene_id VARCHAR, -- systematic like name
+  display_gene_id VARCHAR, -- standard like name
+  species VARCHAR,
+  taxon_id VARCHAR,
+  regulator BOOLEAN,
+  PRIMARY KEY(gene_id, taxon_id)
+); 
+CREATE TABLE spring2022_network.network (
+  regulator_gene_id VARCHAR,
+  target_gene_id VARCHAR,
+  taxon_id VARCHAR,
+  time_stamp TIMESTAMP WITH TIME ZONE,
+  source VARCHAR,
+  FOREIGN KEY (regulator_gene_id, taxon_id) REFERENCES spring2022_network.gene(gene_id, taxon_id),
+  FOREIGN KEY (target_gene_id, taxon_id) REFERENCES spring2022_network.gene(gene_id, taxon_id),
+  FOREIGN KEY (time_stamp, source) REFERENCES spring2022_network.source(time_stamp, source)
+); 

package/database/network-database/scripts/filter_genes.py

@@ -0,0 +1,76 @@
+import psycopg2
+import csv
+import os
+
+PROCESSED_GENES = "../script-results/processed-loader-files/gene.csv"
+MISSING_GENE_DESTINATION = '../script-results/processed-loader-files/missing-genes.csv'
+UPDATE_GENE_DESTINATION = '../script-results/processed-loader-files/update-genes.csv'
+
+try:
+    connection = psycopg2.connect(user="postgres",
+                                  password="",
+                                  host="grnsight2.cfimp3lu6uob.us-west-1.rds.amazonaws.com",
+                                  port="5432",
+                                  database="postgres")
+    cursor = connection.cursor()
+    postgreSQL_select_Query = "select * from spring2022_network.gene"
+
+    cursor.execute(postgreSQL_select_Query)
+    print("Selecting rows from gene table using cursor.fetchall")
+    gene_records = cursor.fetchall()
+
+    db_genes = {}
+    missing_genes = {}
+    genes_to_update = {}
+    for gene in gene_records:
+        # key = (gene_id, taxon_id)
+        key = (gene[0], gene[3])
+        value = {"display_gene_id": gene[1], "species": gene[2], "regulator": gene[4]}
+        db_genes[key] = value
+
+    print(f'Processing file {PROCESSED_GENES}')
+    with open(PROCESSED_GENES, 'r+', encoding="UTF-8") as f:
+        i = 0
+        reader = csv.reader(f)
+        for row in reader:
+            if i != 0:
+                row = row[0].split('\t')
+                gene_id = row[0]
+                display_gene_id = row[1]
+                species = row[2]
+                taxon_id = row[3]
+                regulator = row[4]
+                key = (gene_id, taxon_id)
+                value = {"display_gene_id": display_gene_id , "species": species, "regulator": regulator}
+                if key not in db_genes:
+                    missing_genes[key] = value
+                elif db_genes[key]["display_gene_id"] != display_gene_id:
+                    # the display gene id got updated, so lets update our db to account for that
+                    genes_to_update[key] = value
+            i+=1
+            
+    print(f'Creating missing-genes.csv\n')
+    gene_file = open(MISSING_GENE_DESTINATION, 'w')
+    headers = f'Gene ID\tDisplay Gene ID\tSpecies\tTaxon ID\tRegulator'
+    gene_file.write(f'{headers}\n')
+    for gene in missing_genes:
+        gene_file.write(f'{gene[0]}\t{missing_genes[gene]["display_gene_id"]}\t{missing_genes[gene]["species"]}\t{gene[1]}\t{missing_genes[gene]["regulator"]}\n')
+    gene_file.close()
+
+    print(f'Creating update-genes.csv\n')
+    gene_file = open(UPDATE_GENE_DESTINATION, 'w')
+    headers = f'Gene ID\tDisplay Gene ID\tSpecies\tTaxon ID\tRegulator'
+    gene_file.write(f'{headers}\n')
+    for gene in genes_to_update:
+        gene_file.write(f'{gene[0]}\t{genes_to_update[gene]["display_gene_id"]}\t{genes_to_update[gene]["species"]}\t{gene[1]}\t{genes_to_update[gene]["regulator"]}\n')
+    gene_file.close()
+
+except (Exception, psycopg2.Error) as error:
+    print("Error while fetching data from PostgreSQL", error)
+
+finally:
+    # closing database connection.
+    if connection:
+        cursor.close()
+        connection.close()
+        print("PostgreSQL connection is closed")

package/database/network-database/scripts/generate_network.py

@@ -0,0 +1,185 @@
+from __future__ import print_function
+
+from intermine.webservice import Service
+service = Service("https://yeastmine.yeastgenome.org/yeastmine/service")
+
+import csv
+import re
+import sys
+import os
+import datetime
+
+# Get Network Data from Yeastmine
+
+query = service.new_query("Gene")
+
+query.add_view(
+    "primaryIdentifier", "secondaryIdentifier", "symbol", "name", "sgdAlias",
+    "regulationSummary.summaryParagraph",
+    "regulationSummary.publications.pubMedId",
+    "regulationSummary.publications.citation"
+)
+query.outerjoin("regulationSummary.publications")
+
+regulators = {}
+all_genes = {}
+print("COLLECTING REGULATORS\n")
+for row in query.rows():
+    systematic_name = row["secondaryIdentifier"]
+    standard_name = row["symbol"]
+    if standard_name == None:
+        standard_name = systematic_name
+    
+    regulators[standard_name] =  systematic_name
+    all_genes[standard_name] = systematic_name
+
+regulators_to_targets = {}
+all_targets = {}
+
+
+print("COLLECTING TARGETS\n")
+for regulator in regulators:
+    query = service.new_query("Gene")
+    query.add_constraint("regulatoryRegions", "TFBindingSite")
+    query.add_view(
+        "regulatoryRegions.regulator.symbol",
+        "regulatoryRegions.regulator.secondaryIdentifier", "symbol",
+        "secondaryIdentifier", "regulatoryRegions.regEvidence.ontologyTerm.name",
+        "regulatoryRegions.regEvidence.ontologyTerm.identifier",
+        "regulatoryRegions.experimentCondition",
+        "regulatoryRegions.strainBackground",
+        "regulatoryRegions.regulationDirection",
+        "regulatoryRegions.publications.pubMedId", "regulatoryRegions.datasource",
+        "regulatoryRegions.annotationType"
+    )
+    query.add_sort_order("Gene.secondaryIdentifier", "ASC")
+    query.add_constraint("regulatoryRegions.regulator", "LOOKUP", regulator, "S. cerevisiae", code="A")
+    targets = {}
+
+    for row in query.rows():
+        target_systematic_name = row["secondaryIdentifier"]
+        target_standard_name = row["symbol"]
+        if target_standard_name == None:
+            target_standard_name = target_systematic_name
+        targets[target_standard_name] = target_systematic_name
+        all_targets[target_standard_name] = target_systematic_name
+        all_genes[target_standard_name] =  target_systematic_name
+
+    regulators_to_targets[regulator] = { "systematic_name": regulators[regulator], "targets": targets}
+
+
+
+def create_regulator_to_target_row(target, all_regulators):
+    result = "" + target
+    for regulator in all_regulators:
+        if target in all_regulators[regulator]["targets"]:
+            result += "\t" + "1"
+        else: 
+            result += "\t" + "0"
+    return result
+
+
+# Create files
+
+# Create folder paths 
+if not os.path.exists('../script-results'):
+    os.makedirs('../script-results')
+
+if not os.path.exists('../script-results/networks'):
+    os.makedirs('../script-results/networks')
+
+if not os.path.exists('../script-results/processed-loader-files'):
+    os.makedirs('../script-results/processed-loader-files')
+
+
+
+# Files to be generated
+
+# Generate Networks
+
+REGULATORS_TO_TARGETS_MATRIX = '../script-results/networks/regulators_to_targets.csv'
+REGULATORS_TO_REGULATORS_MATRIX = '../script-results/networks/regulators_to_regulators.csv'
+
+
+targets = []
+for target in all_targets:
+    if target != None:
+        targets.append(target)
+
+regulators_list = []
+for regulator in regulators_to_targets:
+    if regulator != None:
+        regulators_list.append(regulator)
+
+print(f'Creating REGULATORS TO TARGETS MATRIX\n')
+regulator_to_target_file = open(REGULATORS_TO_TARGETS_MATRIX, 'w')
+headers = "cols regulators/rows targets"
+headers += '\t'.join(regulators_list)
+regulator_to_target_file.write(f'{headers}\n')
+for target in targets:
+  result = create_regulator_to_target_row(target, regulators_to_targets)
+  if result != False:
+    regulator_to_target_file.write(f'{result}\n')
+regulator_to_target_file.close()
+
+print(f'Creating REGULATORS TO TARGETS MATRIX\n')
+regulator_to_regulator_file = open(REGULATORS_TO_REGULATORS_MATRIX, 'w')
+headers = "cols regulators/rows targets"
+headers += '\t'.join(regulators_list)
+regulator_to_regulator_file.write(f'{headers}\n')
+for target in targets:
+  result = create_regulator_to_target_row(target, regulators_to_targets)
+  if result != False:
+    regulator_to_regulator_file.write(f'{result}\n')
+regulator_to_regulator_file.close()
+
+
+
+# Create loader-files
+
+# Source Table
+
+SOURCE_DESTINATION = '../script-results/processed-loader-files/source.csv'
+timestamp = datetime.datetime.now(datetime.timezone.utc)
+
+source = "YeastMine - Saccharomyces Genome Database"
+display_name = "Yeastmine - SGD"
+
+source_file = open(SOURCE_DESTINATION, 'w')
+headers = f'Timestamp\tSource\tDisplay Name\n{timestamp}\t{source}\t{display_name}'
+source_file.write(f'{headers}\n')
+source_file.close()
+
+# Gene Table
+
+GENE_DESTINATION = '../script-results/processed-loader-files/gene.csv'
+
+species = "Saccharomyces cerevisiae"
+taxon_id = "559292"
+
+print(f'Creating gene.csv\n')
+gene_file = open(GENE_DESTINATION, 'w')
+headers = f'Gene ID\tDisplay Gene ID\tSpecies\tTaxon ID\tRegulator'
+gene_file.write(f'{headers}\n')
+for gene in all_genes:
+    if gene in regulators:
+        gene_file.write(f'{all_genes[gene]}\t{gene}\t{species}\t{taxon_id}\ttrue\n')
+    else:
+        gene_file.write(f'{all_genes[gene]}\t{gene}\t{species}\t{taxon_id}\tfalse\n')
+
+gene_file.close()
+
+
+# Network Table
+
+NETWORK_DESTINATION = '../script-results/processed-loader-files/network.csv'
+
+
+print(f'Creating network.csv\n')
+network_file = open(NETWORK_DESTINATION, 'w')
+headers = f'Regulator Gene ID\tTarget Gene ID\tTaxon ID\tTimestamp\tSource'
+network_file.write(f'{headers}\n')
+for gene in regulators_to_targets:
+    for target_gene in regulators_to_targets[gene]["targets"]:
+        network_file.write(f'{regulators_to_targets[gene]["systematic_name"]}\t{regulators_to_targets[gene]["targets"][target_gene]}\t{taxon_id}\t{timestamp}\t{source}\n')
+network_file.close()

package/database/network-database/scripts/generate_sgd_network_from_yeastract_network.py

@@ -0,0 +1,120 @@
+from __future__ import print_function
+
+from intermine.webservice import Service
+service = Service("https://yeastmine.yeastgenome.org/yeastmine/service")
+
+import csv
+import os
+import pandas as pd
+
+# Extracting regulator and target genes from file 
+
+# Remember to make the source file folder and put your source files in there
+YEASTRACT_NETWORK = "../source-files/Regulation_matrix_profile2.csv"
+targets = []
+regulators = []
+print(f'Processing file {YEASTRACT_NETWORK}')
+with open(YEASTRACT_NETWORK, 'r+', encoding="UTF-8") as f:
+    targets = []
+    regulators = []
+    i = 0
+    reader = csv.reader(f)
+    for row in reader:
+        if i == 0:
+            # we are getting the regulators
+            j = 0
+            x = row[0].split()
+            for regulator in x:
+                if j > 2:
+                    regulators.append(regulator)
+                j += 1    
+        else:
+            # we are getting the targets
+            targets.append(row[0].split()[0])
+        i+=1
+
+print (targets)
+print (regulators)
+
+regulators_to_targets = {}
+
+print("COLLECTING TARGETS\n")
+for regulator in regulators:
+    query = service.new_query("Gene")
+    query.add_constraint("regulatoryRegions", "TFBindingSite")
+    query.add_view(
+        "regulatoryRegions.regulator.symbol",
+        "regulatoryRegions.regulator.secondaryIdentifier", "symbol",
+        "secondaryIdentifier", "regulatoryRegions.regEvidence.ontologyTerm.name",
+        "regulatoryRegions.regEvidence.ontologyTerm.identifier",
+        "regulatoryRegions.experimentCondition",
+        "regulatoryRegions.strainBackground",
+        "regulatoryRegions.regulationDirection",
+        "regulatoryRegions.publications.pubMedId", "regulatoryRegions.datasource",
+        "regulatoryRegions.annotationType"
+    )
+    query.add_sort_order("Gene.secondaryIdentifier", "ASC")
+    query.add_constraint("regulatoryRegions.regulator", "LOOKUP", regulator, "S. cerevisiae", code="A")
+    regulators_targets = []
+
+    for row in query.rows():
+        target_systematic_name = row["secondaryIdentifier"]
+        target_standard_name = row["symbol"]
+        if target_standard_name == None:
+            target_standard_name = target_systematic_name
+        if target_standard_name in targets:
+            regulators_targets.append(target_standard_name)
+
+    regulators_to_targets[regulator] = regulators_targets
+
+print(regulators_to_targets)
+
+
+
+
+
+
+# We already have the regulator and target genes from Yeastract, now let's find the network
+
+def create_regulator_to_target_row(target, all_regulators):
+    result = "" + target
+    for regulator in all_regulators:
+        if target in all_regulators[regulator]:
+            result += "\t" + "1"
+        else: 
+            result += "\t" + "0"
+    return result
+
+
+# Create files
+
+# Create folder paths 
+if not os.path.exists('../script-results/yeastract-to-sgd-networks'):
+    os.makedirs('../script-results/yeastract-to-sgd-networks')
+
+# Files to be generated
+
+# Generate Networks
+
+SGD_MATRIX = '../script-results/yeastract-to-sgd-networks/SGD_Regulation_matrix_profile2.csv'
+SGD_MATRIX_EXCEL = '../script-results/yeastract-to-sgd-networks/SGD_Regulation_matrix_profile2.xlsx'
+
+print(f'Creating SGD MATRIX\n')
+sgd_matrix_file = open(SGD_MATRIX, 'w')
+headers = "cols regulators/rows targets\t"
+headers += '\t'.join(regulators)
+sgd_matrix_file.write(f'{headers}\n')
+for target in targets:
+  result = create_regulator_to_target_row(target, regulators_to_targets)
+  if result != False:
+    sgd_matrix_file.write(f'{result}\n')
+sgd_matrix_file.close()
+
+# Reading the csv file
+df_new = pd.read_csv(SGD_MATRIX, sep='\t')
+ 
+# saving xlsx file
+GFG = pd.ExcelWriter(SGD_MATRIX_EXCEL)
+df_new.to_excel(GFG, sheet_name="network", index=False)
+ 
+GFG.save()

package/database/network-database/scripts/loader.py

@@ -0,0 +1,79 @@
+import csv
+import re
+# Usage
+# python3 loader.py | psql postgresql://localhost/postgres
+"""
+This program generates direct SQL statements from the source files in order
+to populate a relational database with those files’ data.
+
+By taking the approach of emitting SQL statements directly, we bypass the need to import
+some kind of database library for the loading process, instead passing the statements
+directly into a database command line utility such as `psql`.
+"""
+
+"""
+This function Loads Network Data Sources into the database
+"""
+def LOAD_SOURCES():
+    print('COPY spring2022_network.source (time_stamp, source, display_name) FROM stdin;')
+    NETWORK_DATA_SOURCE = '../script-results/processed-loader-files/source.csv'
+    with open(NETWORK_DATA_SOURCE, 'r+') as f:
+        reader = csv.reader(f)
+        row_num = 0
+        for row in reader:
+            if row_num != 0:
+                r= ','.join(row).split('\t')
+                time_stamp = r[0]
+                source = r[1]
+                display_name = r[2]
+                print(f'{time_stamp}\t{source}\t{display_name}')
+            row_num += 1
+    print('\\.')
+
+
+"""
+This function Loads Gene ID Mapping into the database
+"""
+def LOAD_GENES():
+    print('COPY spring2022_network.gene (gene_id, display_gene_id, species, taxon_id, regulator) FROM stdin;')
+    GENE_SOURCE = '../script-results/processed-loader-files/gene.csv'
+    with open(GENE_SOURCE, 'r+') as f:
+        reader = csv.reader(f)
+        row_num = 0
+        for row in reader:
+            if row_num != 0:
+                r= ','.join(row).split('\t')
+                gene_id = r[0]
+                display_gene_id= r[1]
+                species = r[2]
+                taxon_id = r[3]
+                regulator = r[4]
+                print(f'{gene_id}\t{display_gene_id}\t{species}\t{taxon_id}\t{regulator}')
+            row_num += 1
+    print('\\.')
+
+
+"""
+This function Loads the Network Matrix into the database
+"""
+def LOAD_NETWORK():
+    print('COPY spring2022_network.network (regulator_gene_id, target_gene_id, taxon_id, time_stamp, source) FROM stdin;')
+    NETWORK_SOURCE = '../script-results/processed-loader-files/network.csv'
+    with open(NETWORK_SOURCE, 'r+') as f:
+        reader = csv.reader(f)
+        row_num = 0
+        for row in reader:
+            if row_num != 0:
+                r= ','.join(row).split('\t')
+                regulator_gene_id = r[0]
+                target_gene_id= r[1]
+                taxon_id = r[2]
+                time_stamp = r[3]
+                source = r[4]
+                print(f'{regulator_gene_id}\t{target_gene_id}\t{taxon_id}\t{time_stamp}\t{source}')
+            row_num += 1
+    print('\\.')
+
+LOAD_SOURCES()
+LOAD_GENES()
+LOAD_NETWORK()

package/gdb/PG_VERSION

@@ -0,0 +1 @@
+14