Overview of PIDGINv4
====================

Introduction
------------

Protein target prediction using `Random Forests`_ (RFs) trained on bioactivity data from PubChem_ (extracted Mar 2021) and ChEMBL_ (version 28), using the RDKit_ and Scikit-learn_, which employ a modification of the reliability-density neighbourhood Applicability Domain (AD) analysis by Aniceto [1]_. This project is the sucessor to PIDGIN `version 1`_ [2]_ and PIDGIN `version 2`_ [3]_. This is the updated and retrained version of PIDGIN `version 3`_ Target prediction with extended NCBI pathway and DisGeNET disease enrichment calculation is available as implemented in [4]_.

* Molecular Descriptors : `2048bit RDKit Extended Connectivity FingerPrints`_ (ECFP) [5]_
* Algorithm: `Random Forests`_ with dynamic number of trees (see docs for details), class weight = 'balanced', sample weight = ratio Inactive:Active
* Models generated at four different cut-off's: 100μM, 10μM, 1μM and 0.1μM
* Models generated both with and without mapping to orthologues
* Pathway information from `NCBI BioSystems`_ 
* Disease information from `DisGeNET`_
* Target/pathway/disease enrichment calculated using Fisher's exact test and the Chi-squared test

Details for sizes across all activity cut-off's (needs to be updated for new version):

+------------------------------------------------+-------------------------+---------------------------+
|                                                | Without orthologues     | With orthologues          |
+================================================+=========================+===========================+
| Distinct Models                                | 11,782                  | 16,772                    |
+------------------------------------------------+-------------------------+---------------------------+
| Distinct Targets [exhaustive total]            | 3,698 [11,782]          | 17,021 [63,140]           |
+------------------------------------------------+-------------------------+---------------------------+
| Total Bioactivities Over all models            | 50,210,041              | 437,574,005               |
+------------------------------------------------+-------------------------+---------------------------+
| Actives                                        | 4,079,996               | 4,087,155                 |
+------------------------------------------------+-------------------------+---------------------------+
| Inactives [Of which are Sphere Exclusion (SE)] | 46,130,045 [35,119,663] | 463,237,781 [314,117,438] |
+------------------------------------------------+-------------------------+---------------------------+

Full details on all models are provided in the uniprot_information.txt files in the ortho and no_ortho directories (to be downloaded)


Contributing
------------

Development occurs on GitHub_.
Documentation on Readthedocs_.
Contributions, feature requests, and bug reports are welcome.
Consult the `issue tracker`_.

License
-------
PIDGINv4 is released under the |license_long| (|license|).

Broadly, this means PIDGINv4 can be used in any manner without modification,
with proper attribution. Modification of source code must also be released 
under |license| so that the community may benefit.

Citing PIDGIN
-------------

To cite PIDGINv4, please reference either previous versions [2]_ [3]_ or use |betarelease|.

References
----------

.. [1] |aniceto|
.. [2] |mervin2015|
.. [3] |mervin2018|
.. [4] |mervin2016|
.. [5] |rogers|

.. include:: substitutions.rst