Merge pull request #41 from danielparton/master

1.0.4 release

docs/cli_docs.rst

@@ -42,6 +42,54 @@ Each structure should be named XXX.pdb, where XXX matches the identifier in the
 fasta file. The residues in the coordinate files should also match the
 sequences in the fasta file.
 
+Custom settings
+===============
+
+Many aspects of the behavior of Ensember can be specified by using the Python
+API instead of the main command-line interface.  For API documentation, see the
+source code, or view the docstrings in iPython.
+
+Custom settings via the manual_overrides.yaml file
+--------------------------------------------------
+
+Some options can instead be specified via the ``manual_overrides.yaml`` file,
+which is created when initializing a new Ensembler project. The file contains
+an example configuration, with each line commented out. The user can thus
+uncomment the relevant lines and edit as necessary.
+
+::
+
+  target-selection:
+      domain-spans:
+        ABL1_HUMAN_D0: 242-513
+  template-selection:
+      min-domain-len: 0
+      max-domain-len: 350
+      domain-spans:
+          ABL1_HUMAN_D0: 242-513
+      skip-pdbs:
+          - 4CYJ
+          - 4P41
+          - 4P2W
+          - 4QTD
+          - 4Q2A
+          - 4CTB
+          - 4QOX
+  refinement:
+      ph: 8.0
+      custom_residue_variants:
+          DDR1_HUMAN_D0_PROTONATED:
+              # keyed by 0-based residue index
+              35: ASH
+
+The above configuration makes the following specifications (in order of appearance):
+
+ - Specifies a custom residue span for the target ``ABL1_HUMAN_D0``. This is useful in cases where a different domain span is desired from that annotated in UniProt.
+ - Specifies minimum and maximum domain lengths for templates. Any domain with more than 350 residues would be excluded. The same custom residue span used for target domains is also specified for the template domains.
+ - Certain PDB files can be skipped if they cause problems.
+ - A custom pH level (default: 7.0) is set, which determines how protonation states are assigned by OpenMM prior to molecular dynamics refinement.
+ - Custom residue variants are specified. This can be used to set specific protonation states, rather than rely purely on a defined pH level. These specified protonation states would override those determined by pH. The naming of residue variants (e.g. ``ASH``) follows the OpenMM conventions.
+
 Additional Tools
 ================
 
@@ -64,3 +112,22 @@ The coordinates are simply copied from the first example found for each of
 ``refined-implicit.pdb.gz`` and ``refined-explicit.pdb.gz``. The residue
 numbers are renumbered according to the canonical isoform sequence coordinates
 in the UniProt entry.
+
+Generating unrefined model structures
+-------------------------------------
+
+In some cases it may be useful to analyze model structures which have not
+undergone refinement, but which have topologies equivalent to the final refined
+models. These structures are not saved by the main pipeline functions by
+default, but can be regenerated using
+``ensembler.tools.mktraj.MkTrajImplicitStart``. This code simply loads each
+model structure with ``openmm``, adds hydrogens, and writes the resultant
+structure as a pdb file (``implicit-start.pdb.gz``). It also combines the
+structures into a trajectory (``traj-implicit-start.xtc``). This function is
+accessed via the Python API as follows:
+
+::
+
+  from ensembler.tools.mktraj import MkTrajImplicitStart
+  MkTrajImplicitStart(targetid='EGFR_HUMAN_D0')
+

setup.py

@@ -3,8 +3,8 @@
 from setuptools import setup
 
 ##########################
-VERSION = "1.0.3"
-ISRELEASED = False
+VERSION = "1.0.4"
+ISRELEASED = True
 __version__ = VERSION
 ##########################