docs

README.rst

@@ -58,7 +58,7 @@ Please refer to the `citation page <https://moldrug.readthedocs.io/en/latest/sou
     :target: https://moldrug.readthedocs.io/en/latest/?badge=latest
     :alt: Documentation
 ..  |tests| image:: https://github.com/ale94mleon/moldrug/actions/workflows/conda.yml/badge.svg
-    :target: https://github.com/ale94mleon/moldrug/actions/workflows/conda.yml/
+    :target: https://github.com/ale94mleon/moldrug/actions/workflows/tests.yml/
     :alt: tests
 ..  |pypi-version|  image:: https://img.shields.io/pypi/v/moldrug.svg
     :target: https://pypi.python.org/pypi/moldrug/

docs/source/api.rst

@@ -10,7 +10,7 @@ will be procesed by Vina.
 The idea
 --------
 The idea is simple: we want the best drug for some application. The problem is
-that it is not simple as it sounds. The chemical space is enormous and the estimation
+that it is not as simple as it sounds. The chemical space is enormous and the estimation
 of the fitness is also a term of concern.
 
 Here we address this problem using a GA. The inputs are:
@@ -31,27 +31,40 @@ This cycle will run for ``maxiter`` generations.
 Fitness functions
 -----------------
 
-The default fitness function could be access through::
+The default fitness function could be access through.
 
-```python
-from moldrug import fitness
-cost = fitness.Cost
-```
-A molecule must present several properties to be considered a drug. Some of the must important are: be potent, reach the biological target (good ADME profile) and be real. The last obvious property could be a bottleneck for computer assisted drug design. Because we want to optimize several response variables at the same time; this `cost` function use the concept of desirability functions ([see this paper](https://www.sciencedirect.com/science/article/pii/S0169743911000797)) which optimize several response variables on the hub.
+.. code-block:: python
+
+    from moldrug import fitness
+    cost = fitness.Cost
+
+A molecule must present several properties to be considered a drug. Some of the must important are: 
+be potent, reach the biological target (good ADME profile) and be real. The last obvious property could 
+be a bottleneck for computer assisted drug design. Because we want to optimize several response variables 
+at the same time; this ``cost`` function use the concept of `desirability functions <https://www.sciencedirect.com/science/article/pii/S0169743911000797>`__ 
+which optimize several response variables on the hub.
 
 The following are the response variables:
-1. Vina Scoring Function. [paper](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3041641/)
-1. Quantitative Estimation of Drug-likeness. [paper](https://www.nature.com/articles/nchem.1243)
-2. Synthetic accessibility. [paper](https://jcheminf.biomedcentral.com/articles/10.1186/1758-2946-1-8)
 
-For each of this response variables we create their corresponded Derringer-Suich desirability functions. And them we combine them as a geometric mean:
+#. `Vina Scoring Function. <https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3041641/>`_
+#. `Quantitative Estimation of Drug-likeness. <https://www.nature.com/articles/nchem.1243>`_
+#. `Synthetic accessibility.  <https://jcheminf.biomedcentral.com/articles/10.1186/1758-2946-1-8)>`_
+
+For each of this response variables we create their corresponded Derringer-Suich desirability functions. 
+And them we combine them as a geometric mean:
+
+.. math::
+    D = {\left[\prod_{i = 1}^{3} d_i^{w_i}\right]}^{\frac{1}{\sum_i w_i}}
 
-$$
-D = {\left[\prod_{i = 1}^{3} d_i^{w_i}\right]}^{\frac{1}{\sum_i w_i}}
-$$
 
-where $w_i$ are the weights of each variable; and $d_i$ the desirability functions. Each individual $d_i$ ranges from 0 to 1 and therefore also $D$.
-Because we are looking for the minimum, the function `cost` return $ 1 - D$.
+where :math:`w_i` are the weights of each variable; and :math:`d_i` the desirability functions. 
+Each individual :math:`d_i` ranges from 0 to 1 and therefore also :math:`D`.
+Because we are looking for the minimum, the function `cost` return :math:`1 - D`.
 
-### Multi Receptor
-Could be that our receptor presents high flexibility or that we are interested in generate specific small molecules. In this case could be convenient to add more than one receptor to the cost function. In the `fitness` module the cost function `CostMultiReceptors` tries to reach this goal. For the case of flexibility, we could perform docking in an ensemble of protein structures, and just keep the lower scoring rather that included all of them in the final desirability function.
+Multi Receptor
+--------------
+Could be that our receptor presents high flexibility or that we are interested in generate specific 
+small molecules. In this case could be convenient to add more than one receptor to the cost function. 
+In the ``fitness`` module the cost function ``CostMultiReceptors`` tries to reach this goal. For the case 
+of flexibility, we could perform docking in an ensemble of protein structures, and just keep the lower 
+scoring rather that included all of them in the final desirability function.

docs/source/installation.rst

@@ -12,12 +12,24 @@ Requirements:
 * `CReM <https://github.com/DrrDom/crem>`_ (0.2.9+)
 * `OpenBabel <https://openbabel.org/docs/dev/Installation/install.html>`_ (3.1.0+)
 
+.. note::
+
+    If you have OpenBabel and RDKit already installed you could try with ``conda install moldrug``.
+    But if it is not the case or some version conflicts occurred, think about installed in a isoleated enviroment
+    as it will be show in brief.
+
+
 It is recomendable to install through ``conda``::
 
     $ conda create -n moldrug
     $ conda activate moldrug
     $ conda install -c ale94mleon -c conda-forge -c bioconda moldrug
 
+.. warning::
+
+    Ussually pip has the lates stable version. But we are working to constantlly update the conda packege.
+    Future plans are deployed inside conda-forge.
+
 Another possible way is direclly install from pip. But in this case you must have a correct installation
 of OpenBabel, RDKit and autodock-vina. One posibility is::
 
@@ -41,4 +53,5 @@ or::
     # To get the last "stable" version. This project is still in beta state.
     pip install moldrug
     
-We are currently working in a ``docker`` container.
+We are currently working in a ``docker`` container. But you could use the `Docker configuration file on GitHub <https://github.com/ale94mleon/MolDrug/blob/main/Dockerfile>`__. 
+and ``pip install moldrug`` inside it.

docs/tmp/packaging.md

@@ -9,3 +9,10 @@ conda build -c conda-forge -c bioconda moldrug
 ```
 
 Valid `anaconda upload` command, with correct paths, is printed in the terminal at the end of building.
+
+## For the documentation
+
+```bash
+make html
+sphinx-build -b html source/. public
+```