miga-base 1.2.18.1 → 1.3.0.0

data/utils/FastAAI/fastaai/fastaai_miga_crystals_to_db.py

@@ -0,0 +1,297 @@
+import sys
+import os
+
+import gzip
+import argparse 
+import json
+import gzip
+
+import numpy as np
+
+import multiprocessing
+import sqlite3
+
+def convert_array(bytestring):
+	return np.frombuffer(bytestring, dtype = np.int32)
+
+sqlite3.register_converter("array", convert_array)
+
+class acc_indexer:
+	def __init__(self):
+		self.forward = None
+		self.reverse = None
+		self.generate_accessions_index()
+
+	def list_to_index_dict(self, list):
+		result = {}
+		counter = 0
+		for item in list:
+			result[item] = counter
+			counter += 1
+		return result
+		
+	def rev_list_to_index_dict(self, list):
+		result = {}
+		counter = 0
+		for item in list:
+			result[counter] = item
+			counter += 1
+		return result
+		
+	def generate_accessions_index(self):
+		acc_list = ['PF01780_19', 'PF03948_14', 'PF17144_4', 'PF00830_19', 'PF00347_23', 'PF16906_5', 'PF13393_6',
+			'PF02565_15', 'PF01991_18', 'PF01984_20', 'PF00861_22', 'PF13656_6', 'PF00368_18', 'PF01142_18', 'PF00312_22', 'PF02367_17',
+			'PF01951_16', 'PF00749_21', 'PF01655_18', 'PF00318_20', 'PF01813_17', 'PF01649_18', 'PF01025_19', 'PF00380_19', 'PF01282_19',
+			'PF01864_17', 'PF01783_23', 'PF01808_18', 'PF01982_16', 'PF01715_17', 'PF00213_18', 'PF00119_20', 'PF00573_22', 'PF01981_16',
+			'PF00281_19', 'PF00584_20', 'PF00825_18', 'PF00406_22', 'PF00177_21', 'PF01192_22', 'PF05833_11', 'PF02699_15', 'PF01016_19',
+			'PF01765_19', 'PF00453_18', 'PF01193_24', 'PF05221_17', 'PF00231_19', 'PF00416_22', 'PF02033_18', 'PF01668_18', 'PF00886_19',
+			'PF00252_18', 'PF00572_18', 'PF00366_20', 'PF04104_14', 'PF04919_12', 'PF01912_18', 'PF00276_20', 'PF00203_21', 'PF00889_19',
+			'PF02996_17', 'PF00121_18', 'PF01990_17', 'PF00344_20', 'PF00297_22', 'PF01196_19', 'PF01194_17', 'PF01725_16', 'PF00750_19',
+			'PF00338_22', 'PF00238_19', 'PF01200_18', 'PF00162_19', 'PF00181_23', 'PF01866_17', 'PF00709_21', 'PF02006_16', 'PF00164_25',
+			'PF00237_19', 'PF01139_17', 'PF01351_18', 'PF04010_13', 'PF06093_13', 'PF00828_19', 'PF02410_15', 'PF01176_19', 'PF02130_17',
+			'PF01948_18', 'PF01195_19', 'PF01746_21', 'PF01667_17', 'PF03874_16', 'PF01090_19', 'PF01198_19', 'PF01250_17', 'PF17136_4',
+			'PF06026_14', 'PF03652_15', 'PF04019_12', 'PF01201_22', 'PF00832_20', 'PF01264_21', 'PF03840_14', 'PF00831_23', 'PF00189_20',
+			'PF02601_15', 'PF01496_19', 'PF00411_19', 'PF00334_19', 'PF00687_21', 'PF01157_18', 'PF01245_20', 'PF01994_16', 'PF01632_19',
+			'PF00827_17', 'PF01015_18', 'PF00829_21', 'PF00410_19', 'PF00833_18', 'PF00935_19', 'PF01992_16']
+		
+		self.forward = list_of_poss_accs = self.list_to_index_dict(acc_list)
+		self.reverse = list_of_poss_accs = self.rev_list_to_index_dict(acc_list)
+	
+class miga_db_builder:
+	def __init__(self, path):
+		self.path = path
+		self.conn = None
+		self.curs = None
+		
+	def activate(self):
+		self.conn = sqlite3.connect(self.path)
+		self.curs = self.conn.cursor()
+		
+	def deactivate(self):
+		self.curs.close()
+		self.conn.close()
+		self.conn = None
+		self.curs = None	
+		
+	def initialize_metadata(self):
+		self.curs.execute("CREATE TABLE IF NOT EXISTS genome_index (genome text, gen_id integer, protein_count integer)")
+		self.curs.execute("CREATE TABLE IF NOT EXISTS genome_acc_kmer_counts (genome integer, accession integer, count integer)")
+		self.curs.execute("CREATE INDEX IF NOT EXISTS kmer_acc ON genome_acc_kmer_counts (genome, accession);")
+		
+	def insert_genome_index(self, gi):
+		self.curs.executemany("INSERT INTO genome_index VALUES (?, ?, ?)", gi)
+		self.conn.commit()
+		
+	def insert_gak(self, gak):
+		self.curs.executemany("INSERT INTO genome_acc_kmer_counts VALUES (?, ?, ?)", gak)
+		self.conn.commit()
+		
+	def add_acc_genomes(self, acc, data):
+		create_sql = "CREATE TABLE IF NOT EXISTS {acc}_genomes (genome INTEGER PRIMARY KEY, kmers array)"
+		create_sql = create_sql.format(acc = acc)
+		self.curs.execute(create_sql)
+		insert_sql = "INSERT INTO {acc}_genomes VALUES (?, ?)"
+		insert_sql = insert_sql.format(acc=acc)
+		self.curs.executemany(insert_sql, data)
+		
+		self.conn.commit()
+		
+		
+	def add_acc_kmers(self, acc, data):
+		create_sql = "CREATE TABLE IF NOT EXISTS {acc} (kmer INTEGER PRIMARY KEY, genomes array)"
+		create_sql = create_sql.format(acc = acc)
+		self.curs.execute(create_sql)
+		insert_sql = "INSERT INTO {acc} VALUES (?, ?)"
+		insert_sql = insert_sql.format(acc=acc)
+		self.curs.executemany(insert_sql, data)
+		
+		self.conn.commit()
+		
+		self.curs.execute("CREATE INDEX {acc}_index ON {acc} (kmer)".format(acc=acc))
+		self.conn.commit()
+	
+#Class for loading crystal files and prepping them for consumption.
+class ravenous_crystal_lizard:
+	def __init__(self, crystal_list, database, overwrite = False):
+		self.paths_file = crystal_list
+		self.input_paths = None
+		
+		self.crystal_contents = None
+		
+		self.accession_index = acc_indexer()
+		
+		self.genome_index = None
+		self.genome_prot_ct = None
+		self.gak = None
+		
+		self.db_already_exists = os.path.exists(database)
+		self.overwrite = overwrite
+		self.og_db_path = database
+		self.db = miga_db_builder(database)
+		
+	def consume_list(self):
+		with open(self.paths_file) as fh:
+			self.input_paths = fh.readlines()
+			
+		self.input_paths = [path.strip() for path in self.input_paths]
+				
+	def consume_crystal_data(self):
+		self.crystal_contents = {}
+		self.genome_index = []
+		self.gak = []
+		current_index = 0
+		for crystal in self.input_paths:
+			if crystal.endswith(".gz"):
+				with gzip.open(crystal, "rb") as fh:
+					next_crystal = fh.read()
+					next_crystal = next_crystal.decode('utf-8')
+					next_crystal = json.loads(next_crystal)
+			else:
+				with open(crystal, "r") as fh:
+					next_crystal = json.load(fh)
+				
+			filename = next_crystal["filename"]
+			next_crystal = next_crystal["protein_data"]
+			protein_count = len(next_crystal)
+			
+			next_index = (filename, current_index, protein_count,)
+			self.genome_index.append(next_index)
+			
+			for acc in next_crystal:
+				acc_id = self.accession_index.forward[acc]
+				
+				if acc not in self.crystal_contents:
+					self.crystal_contents[acc] = {}
+				
+				kmer_list = np.array(next_crystal[acc]["kmers"], dtype = np.int32)
+				kmer_ct = kmer_list.shape[0]
+				
+				next_gak = (current_index, acc_id, kmer_ct, )
+				self.gak.append(next_gak)
+				
+				self.crystal_contents[acc][current_index] = kmer_list
+				
+			current_index += 1
+					
+		self.db.activate()
+		self.db.initialize_metadata()
+		self.db.insert_genome_index(self.genome_index)
+		self.db.insert_gak(self.gak)
+		
+		for acc in self.crystal_contents:
+			#self.db.add_acc_genomes(acc, self.crystal_contents[acc])
+			flipped_dataset = self.invert_to_kmer_first(self.crystal_contents[acc])
+			self.db.add_acc_kmers(acc, flipped_dataset)
+			flipped_dataset = None
+			insertable_genomes = []
+			for genome, kmer_array in self.crystal_contents[acc].items():
+				next_row = (genome, kmer_array.tobytes(),)
+				insertable_genomes.append(next_row)
+				
+			self.db.add_acc_genomes(acc, insertable_genomes)
+			self.crystal_contents[acc] = None
+					
+		self.db.deactivate()
+		
+	#Take a set of genome : kmer_lists and flip them to an equivalent set of kmer : genome_lists
+	def invert_to_kmer_first(self, dataset):
+		genomes = []
+		counts = []
+		kmer_unlist = []
+		for genome_index in dataset:
+			genomes.append(genome_index)
+			counts.append(dataset[genome_index].shape[0])
+			kmer_unlist.append(dataset[genome_index])
+		
+		genomes = np.array(genomes, dtype = np.int32)
+		counts = np.array(counts, dtype = np.int32)
+		
+		kmer_unlist = np.concatenate(kmer_unlist) #A 1-d array of all of the kmers for all of the genomes containing this SCP
+		counted_gens = np.repeat(genomes, counts) #A 1-d array of the same length as kmer_unlist with the corresp. genome index for each kmer
+		
+		#This contains a list of kmers and genome indices repeated enough times to match their kmer collection 1 to 1 in the same order
+		formatted_pairs = np.vstack([kmer_unlist, counted_gens]) 
+		kmer_unlist = None
+		counted_gens = None
+				
+		#Sort the list based on kmer, then genome
+		sorted_indices = np.lexsort((formatted_pairs[1, :], formatted_pairs[0, :]))
+		
+		formatted_pairs = formatted_pairs[:, sorted_indices]
+
+		#Collect an ordered list of unique kmers
+		discovered_kmers = np.unique(formatted_pairs[0, :])
+		
+		#Collect a list of the genomes associated with each kmer
+		formatted_pairs = np.split(formatted_pairs[1, :], np.unique(formatted_pairs[0, :], return_index=True)[1][1:])
+
+		final_dataset = []
+		for kmer, genomes in zip(discovered_kmers, formatted_pairs):
+			genome_bytestring = genomes.tobytes()
+			
+			kmer = int(kmer)
+			
+			final_dataset.append((kmer, genome_bytestring,))
+			
+		return final_dataset
+		
+	def run(self):
+		do_run = True
+		if self.db_already_exists:
+			if self.overwrite:
+				os.remove(self.og_db_path)
+			else:
+				print("")
+				print("Target database file already exists! I'm quitting.")
+				print("Supply a different path or use --overwrite")
+				do_run = False
+				
+		if do_run:
+			self.consume_list()
+			self.consume_crystal_data()
+		
+#Add options	
+def options():
+	parser = argparse.ArgumentParser(formatter_class=argparse.RawTextHelpFormatter,
+			description='''Dedicated MiGA db builder module.
+			
+	Takes a file containing a list of paths to crystals and builds a database from those.
+			
+	Notes:
+			
+	Assumes the supplied database path does not exist. 
+	Use --overwrite to delete an existing DB under the same path if you want to.''')
+			
+	parser.add_argument('--crystal_list',   dest = 'crystals', default = None, help = 'File containing a list of paths to FastAAI crystals')
+	parser.add_argument('--database_path',   dest = 'db', default = None, help = 'Path to a NEW database to be built.')
+	
+	parser.add_argument('--overwrite',   dest = 'overwrite', action = 'store_true', help = 'Delete an existing database at --database_path and create a new one. Otw. quits to preserve existing db.')
+		
+	args, unknown_opts = parser.parse_known_args()
+	
+	return parser, args
+
+def main():
+	p, a = options()
+	crystal_file = a.crystals
+	db = a.db
+	overwrite = a.overwrite
+	
+	if len(sys.argv) < 3:
+		p.print_help()
+	
+	if crystal_file is None:
+		print("I need a file containing a list of paths to crystals")
+		sys.exit()
+		
+	if db is None:
+		print("I need a path to an output database")
+		sys.exit()
+	
+	mn = ravenous_crystal_lizard(crystal_list = crystal_file, 
+								database = db, 
+								overwrite = overwrite)
+	mn.run()
+		
+if __name__ == "__main__":
+	main()