comparison tools/protein_analysis/tmhmm2.py @ 1:3ff1dcbb9440

Migrated tool version 0.0.3 from old tool shed archive to new tool shed repository

0:bca9bc7fdaef

The second major potential feature is taking advantage of multiple cores
(since TMHMM v2.0 itself is single threaded) by dividing the input FASTA file
into chunks and running multiple copies of TMHMM in parallel. I would normally
use Python's multiprocessing library in this situation but it requires at
least Python 2.6 and at the time of writing Galaxy still supports Python 2.4.
"""
import sys
import os
from seq_analysis_utils import stop_err, split_fasta, run_jobs

   stop_err("Threads argument %s is not a positive integer" % sys.argv[1])
fasta_file = sys.argv[2]
tabular_file = sys.argv[3]

def clean_tabular(raw_handle, out_handle):
    """Clean up tabular TMHMM output."""
    for line in raw_handle:
        if not line:
            continue
        parts = line.rstrip("\r\n").split("\t")
        try:

        first60 = first60[8:]
        assert predhel.startswith("PredHel="), line
        predhel = predhel[8:]
        assert topology.startswith("Topology="), line
        topology = topology[9:]
	out_handle.write("%s\t%s\t%s\t%s\t%s\t%s\n" \
                   % (identifier, length, expAA, first60, predhel, topology))

fasta_files = split_fasta(fasta_file, tabular_file, FASTA_CHUNK)
temp_files = [f+".out" for f in fasta_files]
jobs = ["tmhmm %s > %s" % (fasta, temp)
        for fasta, temp in zip(fasta_files, temp_files)]

out_handle = open(tabular_file, "w")
out_handle.write("#ID\tlen\tExpAA\tFirst60\tPredHel\tTopology\n")
for temp in temp_files:
    data_handle = open(temp)
    clean_tabular(data_handle, out_handle)
    data_handle.close()
out_handle.close()

clean_up(fasta_files)
clean_up(temp_files)

1:3ff1dcbb9440

The second major potential feature is taking advantage of multiple cores
(since TMHMM v2.0 itself is single threaded) by dividing the input FASTA file
into chunks and running multiple copies of TMHMM in parallel. I would normally
use Python's multiprocessing library in this situation but it requires at
least Python 2.6 and at the time of writing Galaxy still supports Python 2.4.

Also tmhmm2 can fail without returning an error code, for example if run on a
64 bit machine with only the 32 bit binaries installed. This script will spot
when there is no output from tmhmm2, and raise an error.
"""
import sys
import os
from seq_analysis_utils import stop_err, split_fasta, run_jobs

   stop_err("Threads argument %s is not a positive integer" % sys.argv[1])
fasta_file = sys.argv[2]
tabular_file = sys.argv[3]

def clean_tabular(raw_handle, out_handle):
    """Clean up tabular TMHMM output, returns output line count."""
    count = 0
    for line in raw_handle:
        if not line:
            continue
        parts = line.rstrip("\r\n").split("\t")
        try:

        first60 = first60[8:]
        assert predhel.startswith("PredHel="), line
        predhel = predhel[8:]
        assert topology.startswith("Topology="), line
        topology = topology[9:]
        out_handle.write("%s\t%s\t%s\t%s\t%s\t%s\n" \
                   % (identifier, length, expAA, first60, predhel, topology))
        count += 1
    return count

#Note that if the input FASTA file contains no sequences,
#split_fasta returns an empty list (i.e. zero temp files).
fasta_files = split_fasta(fasta_file, tabular_file, FASTA_CHUNK)
temp_files = [f+".out" for f in fasta_files]
jobs = ["tmhmm %s > %s" % (fasta, temp)
        for fasta, temp in zip(fasta_files, temp_files)]

out_handle = open(tabular_file, "w")
out_handle.write("#ID\tlen\tExpAA\tFirst60\tPredHel\tTopology\n")
for temp in temp_files:
    data_handle = open(temp)
    count = clean_tabular(data_handle, out_handle)
    data_handle.close()
    if not count:
        clean_up(fasta_files)
        clean_up(temp_files)
        stop_err("No output from tmhmm2")
out_handle.close()

clean_up(fasta_files)
clean_up(temp_files)