To Do List

To Do in Biskit.ChainSeparator

Use biopython insted of Blast2Seq, making the latter obsolete
Create an override for the chain comparison if one wants to keep identical chains (i.e homodimers)

To Do in Biskit.Dock.ComplexEvolvingList.ComplexEvolvingList

implement plotting functions for evolving Complexes
right now normal Complexes are tolerated
adapt model management

To Do in Biskit.Dock.ComplexList.ComplexList

Removing items with pop(), del, remove() etc. will not remove unused PDBModels from rec_models or lig_models.

To Do in Biskit.Dock.Docker.Docker

implement that as JobMaster / JobSlave instead, so that also the Hex-file parsing is distributed.

To Do in Biskit.Dock.Docker.RunThread

implement that as JobSlave instead

To Do in Biskit.EnsembleTraj.EnsembleTraj.concatEnsembles

fix so that pc, and profiles are not lost

To Do in Biskit.EnsembleTraj.EnsembleTraj.memberList

cast member Trajectories back to normal Trajectory object

To Do in Biskit.EnsembleTraj.EnsembleTraj.takeMembers

return self.__class__ instead of EnsembleTraj

To Do in Biskit.IcmCad.IcmCad

Not speed-optimized!
- it should be possible to pipe the PDB-files directly into icmbrowser instead of writing temporary files to disc
- Abagyan and Totrov mention a c-library available -- the best solution would hence be to wrap this c-library in python...

To Do in Biskit.LocalPath.LocalPath

input: allow multiple or overlapping substitutions output: only substitute what is necessary until path exists

To Do in Biskit.Mod.SequenceSearcher.SequenceSearcher

copy blast output

To Do in Biskit.Mod.SequenceSearcher.SequenceSearcher.remoteBlast

Remote Blasting is running as expected but sequences are still retrieved from a local database. Implement remote collection of fasta seuqences from NCBI (there should be something like in Biopython). Otherwise something like this will also work:

 ## collect the entry with gi 87047648
 url = 'http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?CMD=text&db=protein&dopt=FASTA&dispmax=20&uid=87047648'
 import urllib
 handle = urllib.urlopen(url)
 lines = handle.readlines()
 seq = ''
 for i in range(len(lines)):
     if re.match( "[<]pre[>][<]div class='recordbody'[>]{2}gi", l[i] ):
         i+= 1
         while not re.match( "^[<]/pre[>][<]/div[>]", l[i]):
             seq += l[i][:-1]
             i+= 1
 print seq

To Do in Biskit.Mod.SequenceSearcher.SequenceSearcher.remoteBlast

 ## collect the entry with gi 87047648
 url = 'http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?CMD=text&db=protein&dopt=FASTA&dispmax=20&uid=87047648'
 import urllib
 handle = urllib.urlopen(url)
 lines = handle.readlines()
 seq = ''
 for i in range(len(lines)):
     if re.match( "[<]pre[>][<]div class='recordbody'[>]{2}gi", l[i] ):
         i+= 1
         while not re.match( "^[<]/pre[>][<]/div[>]", l[i]):
             seq += l[i][:-1]
             i+= 1
 print seq

To Do in Biskit.PDBModel.PDBModel

clean up, rename returnPdb()
clean up the __terAtoms mess
use Profiles instead of space consuming atom dictionaries

To Do in Biskit.PDBModel.PDBModel.equalAtoms

option to make sure atoms are also in same order

To Do in Biskit.PDBModel.PDBModel.equalAtoms

option to make sure atoms are also in same order

To Do in Biskit.PVM.TrackingJobMaster.TrackingJobMaster

try finding some solution to the problem where the master sends out an exit signal to all slaves but doesn't wait for a response (see note)
test restart function
restart data are not automatically saved (e.g. in intervals)

To Do in Biskit.Xplorer.Xplorer

Most of this has been abstracted into Executor next thing to do is to subclass Executor and delete redundant parts