This is a read-only mirror of pymolwiki.org
User talk:Philaltist
Revision as of 17:18, 9 March 2005 by Philaltist (talk | contribs)
Hi everyone !
I have created a script because I was tired of browsing the pdb for endless searches of structures relevant to my field. This scripts takes as input a text file in which you copied your favourite keywords, and the number of days you want to search back (by default, it will look at the structures added the last 50 days). It generates a report text file that contains the pdb id and name of the relevant structures that have been recently added. All you need to do is add 'run pdbsurvey.py' to your startup file, and create a text file called 'keywords.txt' with your keywords separated by an end-of-line character. Then you're ready to go. Just hit 'pdbsurvey' from within your PyMol instance, and the program returns the report file. The pdb ftp server is updated every tuesday.
from pymol import cmd
def pdbsurvey(days=50):
"""USAGE : pdbsurvey (<days>)
Surveys the updates added to the PDB (ftp.rcsb.org) in the last
50 days (or otherwise specified when calling this function) for
entries that contain the words specified in the file
keywords.txt.
"""
print days
import ftplib
import time
import os
import string
def todaymerge():
"""Puts today's date in a pdb format string.
"""
date=time.localtime()
fyear="%i" %(date[0])
fmonth="%i" %(date[1])
if date[1]<10:
fmonth="0"+"%i" %(date[1])
fday="%i" %(date[2])
if date[2]<10:
fday="0"+"%i" %(date[2])
dateS=fyear+fmonth+fday
return dateS
def file2list(filename):
"""Low-level routine to brainlessly implement
file.read().
"""
fq=open(filename,'rb')
linesS=fq.read()
fq.close()
LIST=linesS.splitlines()
return LIST
def connect2pdb():
"""Opens an anonymous socket to ftp://ftp.rcsb.org
"""
f=ftplib.FTP()
f.connect ('ftp.rcsb.org')
f.login ()
print "Remote connection established","\n"
return f
def decrementdate(dateS):
"""given a string date (pdb format yyyymmdd)
this routine returns a string of the day before
(sadly assuming that every month has 31 days, but
no big deal here...).
"""
#decompose dateS into components
yearS=dateS[0]+dateS[1]+dateS[2]+dateS[3]
monthS=dateS[4]+dateS[5]
dayS=dateS[6]+dateS[7]
#convert each into integers
yearI=int(yearS)
monthI=int(monthS)
dayI=int(dayS)
#actual stuff
dayI=dayI-1
if dayI==0:
dayI=31
monthI=monthI-1
if monthI==0:
monthI=12
yearI=yearI-1
dayS="%i" %(dayI)
monthS="%i" %(monthI)
yearS="%i" %(yearI)
if dayI<10:
dayS="0"+dayS
if monthI<10:
monthS="0"+monthS
#and finally...
dateS=yearS+monthS+dayS
return dateS
def findlastdir(dateS,f,days):
"""Puts the names of the "recent" directories in the
list named "directoriesL".
"""
directoriesL=['']
j=p=0
while p<days:
dateS=decrementdate(dateS)
attempt="/pub/pdb/data/status/"+dateS
try :
f.cwd(attempt)
directoriesL[j:j]=[attempt]
j=j+1
except :
pass
p=p+1
directoriesL.pop()
return directoriesL
def compilinfile(directoriesL,f):
"""lists all structures in the added.pdb files
contained in the directories specified in directoriesL
"""
command="RETR added.pdb"
handle=open("donotedit.dat","wrb")
for k in directoriesL:
f.cwd(k)
print "Currently in directory ",f.pwd()
f.retrbinary(command,handle.write)
handle.close()
return len(directoriesL)
def listparser():
"""Extracts the pdbids from donotedit.dat file,
and stacks them into the list pdbidsL
"""
linesL=file2list("donotedit.dat")
pdbidsL=[]
for iter in linesL:
pdbidsL.append(iter[57:61])
for iter in pdbidsL:
iter=string.lower(iter)
pdbidsL.sort()
return pdbidsL
def currentrelease(f):
"""Stores the content of cmpd_res.idx file
This file contains the equivalencies pdbid<->title
for all current entries of the PDB.
"""
command="RETR cmpd_res.idx"
f.cwd("/pub/pdb/derived_data/index/")
print "Currently in directory ",f.pwd()
fq=open("dictionnary.dat",'wrb')
f.retrbinary(command,fq.write)
fq.close()
dictL=file2list("dictionnary.dat")
return dictL
def extract(pdbidsL,dictL):
"""Populates dictionnaryD with pdb entries found in the
latest releases.
"""
dictionnaryD={}
problemL=[]
extractL=[dictionnaryD,problemL]
for i in dictL:
tempS=string.lower(i[0:4])
for ii in pdbidsL:
if ii == tempS:
title=i[14:216]
extractL[0][ii]=title
if len(extractL[0].keys()) != len(pdbidsL):
print "Dimension mismatch, seeking troublemaker..."
for i in pdbidsL:
equiv=0
for ii in extractL[0].keys():
if i==ii:
equiv=equiv+1
if equiv==0:
extractL[1].append(i)
return extractL
def disconnectpdb(f):
"""Diconnects the current ftp session
"""
f.quit()
print "Remote connection terminated","\n"
return f
def releventries(dictionnaryD):
"""Generates a cleaned dictionnary with only entries
that have one or more keywords specified in the local
user-defined keywords.txt file
"""
keywL=file2list("keywords.txt")
relevdicD={}
for i in keywL:
for elem in dictionnaryD.keys():
temp=dictionnaryD[elem]
if temp.find(i) != -1:
relevdicD[elem]=temp
return relevdicD
def diskcleanup(filelist=["donotedit.dat","dictionnary.dat"]):
"""Lo-level disk cleanup to free up memory without the user
"""
for filename in filelist:
command='DEL '+filename
os.system(command)
return "clean"
print "Welcome in the auto-PDB updater !"
print "Survey of updates made since",days,"days ago."
print "Acquisition of local time..."
dateS=todaymerge() #Initializes dateS
print "today is ",dateS
print "Connecting to remote ftp server..."
f=connect2pdb() #Connect anonymously to ftp.rcsb.org
print "Acquisition of latest added remote directories..."
directoriesL=findlastdir(dateS,f,days) #Lists recent directories in directoriesL
if len(directoriesL)==0:
print "No updates have been found since",days,"ago. Starting over with 50 days ago."
directoriesL=findlastdir(dateS,f,50)
print "Acquisition of latest addedremote files..."
updatesnumberI=compilinfile(directoriesL,f) #Concatenates the corresponding added.pdb into donotedit.dat
print "Parsing of latest entries..."
pdbidsL=listparser() #Recent names now present in the pdbidsL list (one name per element)
print "Acquisition of the current pdb distribution..."
dictL=currentrelease(f) #Populates dictL with the current entries of the PDB
print "Parsing of the current pdb distribution into [code,title] tuples..."
extractL=extract(pdbidsL,dictL) #generates the dictionnary of latest releases key:PDBid ; definition:pdbtitle
print "Disconnection from the remote ftp server..."
f=disconnectpdb(f) #Closes the ftp instance
print "Extraction of the relevant entries..."
relevdicD=releventries(extractL[0]) #Generates a subset of dictionnary D with criterion being "has keywords contained in keywords.txt in its title"
print "Cleaning program-generated temporary files..."
clean=diskcleanup() #Cleans the mess generated by the program
reportL=[]
reportL.append("\n")
reportL.append("###############REPORT########################################\n")
reportL.append("\n")
lendictS="%i" %(len(dictL))
chmilblik = 'The current pdb version (as of '+dateS+") has "+lendictS+" entries.\n"
reportL.append(chmilblik)
line="The most recent directory is : "+directoriesL[0]+".\n"
reportL.append(line)
updatesnumberS="%i" %(updatesnumberI)
entriesnumber="%i" %(len(extractL[0].keys()))
line="The "+updatesnumberS+" last updates ("+entriesnumber+" entries) have been examined.\n"
reportL.append(line)
diclengthS="%i" %(len(relevdicD.keys()))
line=diclengthS+" are relevant to you :\n"
reportL.append(line)
for i in relevdicD.keys():
entry=i+" : "+relevdicD[i]+"\n"
reportL.append(entry)
problemS=""
for i in extractL[1]:
problemS=i+";"+problemS
problemS="["+problemS
problemS=problemS.strip(";")
problemS=problemS+"]"
lineS="The entries "+problemS+" raised problems,"
reportL.append(lineS)
reportL.append("they should be examined manually.")
reportL.append("\n")
reportL.append("###############END OF REPORT#################################\n")
report=open("report.aut","w")
for elem in reportL:
print elem
elem=elem+'\n'
report.writelines(elem)
report.close()
command2='start keywords.txt'
command3='start report.aut'
os.system(command2)
os.system(command3)
cmd.extend("pdbsurvey",pdbsurvey)
Thank you for any feedback, ways of improving it,...