Find file
Fetching contributors…
Cannot retrieve contributors at this time
38 lines (21 sloc) 2.67 KB
- Improved parsing of whitespace for input sequence
- Added and ran additional tests
= Bug fixes
- minor/various
* The kappa-P (proline/charge patterning) parameter has been added. This is used in the upcoming paper by Martin and Holehouse (reference to come).
* Three different sequence complexity measures have been added: Wooton-Federhen complexity [8], Linguistic complexity [9], Lempel-Ziv-Welch (LZW) complexity [10]. These algorithms have a number of parameters which can be defined, as well as the ability to use a set of pre-defined reduced alphabetes [11]
* Overall PPII propensity can be calculated based on the amino acid PPII propensity as defined by Elam *et al.* [6].
* All figures can now be saved as PDFs (useful for post-process with graphics tools, e.g. Adobe Illustrator)
* The bar positions for blob-based figures have been updated such that bin centers are used on all figures. This makes comparing figures using different blob-widths much easier.
* Numpy vectors of the linear NCPR, FCR, hydropathy and sequence complexity can now be directly returned for further analysis
* Sequences can be directly converted to a reduced alphabet sequence. The mapping of true amino-acid alphabet to reduced alphabet can take advantage of 11 predefined mappings, or this mapping can be defined using a custom dictionary.
* localCIDER can now run under Python 3
* Local compositional analysis has been added to help identify local regions which are enriched in a particular type of amino acid. Proteins are complex heteropolymers, so plotting the local density of each amino acid with a sliding window is very difficult to visualize and interpret. To overcome this challenge, we take a two-pronged approach: firstly, amino acids are group by physiochemical properties. Secondly, we then fit the local density noisy data to a univariate spline fit to give a smoothed description of the local density along the sequence. Combined, this allows us to generate easy-to-read plots that illustrate the local density of residues at any given location along the sequence
* The Omega patterning parameter (previously kappa_proline) is re-defined as per the manuscript by Martin & Holehouse *et al.*
* A generalized patterning parameter for considering 2 or 3 letter alphabets is provided (kappa_X), allowing users to ask general questions about sequence patterning.
* The figure generation code has been further optimised, as has code documentation
If there are sequence feautures you would like to see added to localCIDER please don't hesitate to get in touch - we're always looking for new features to add and to further grow localCIDER as a general purpose protein analysis framework.