The pipeline contains many parameters and functions, most of which are possible to modify. Since the source code is available in its entirety it is most certainly possible to modify it to your heart’s content. The source is very well commented, making it possible for anyone comfortable with Python to modify any aspect of the pipeline to their liking.
Here are some examples of modifications that might interest some users.
The classification function is very easy to modify. The first a user should try is to modify the classification function parameters; K, M, C, D. Using these four parameters, the function’s shape can be modified in many ways. Most probably the user is interested in changing the two-part linear classification function to a regular single score cutoff. This is easy to do by just setting D to 0 and C to whatever cutoff you are interested in.
To modify the actual function, the user should open qnrpipeline.py and modify the function definition on line 421:
classificationfunction = lambda L: options.classifyK*L + options.classifyM
This function can be modified to whatever function the user is interested in, but if it you need it to take other arguments than fragment length, you have to modify the classification function in module fluff.py. This is non-trivial, please refer to its specific implementation if you need this kind of very advanced functionality.
The user can supply the path to a FASTA file containing a couple of reference sequences to which all clusters will be aligned. This can be specified by supplying the --alignseqpath command line option. This command and its related functions have many hardcoded parameters specified for PMQR gene variants and to modify this the user has to go into the related functions in fluff.py.
In its current state it requires that the supplied FASTA file contains PMQR sequences with the labels “QnrA”, “QnrB”, “QnrC”, “QnrD”, or “QnrS” in the headers. For simplicity, do not supply a file containing more than a few sequences (for example QnrA1, QnrB1, QnrC, QnrD, QnrS1).