The rase-rna-structural-stability-estimator from fabriziocosta

Mult processing support

Do we already have plfold step parallelized? That would be good to support this to have 24core support on the cluster. If it is already working in parallel we need to expose n-cores option to the user interface.

output file argument

please add a -o,--outfile argument to write the STDOUT output directly to file...

make plot file format an option

svg
pdf
png

layouting enhancement?

add diagonal invisible edges of sqrt(edgelength) into stacking faces
add position information (every 10)
within individual mutation plots: highlight the mutation (2-letter label + vertex (boundary) coloring)

heatmap plot

raw file (col header=(ACGU), row names = positions (starting with 1))
horizontal heatmap plot (x-axis starts with 1) y-axis annotated with ACGU (lex ordered) at both ends

Graph Context length option

To get a (super) local comparison score between wild and mutant feature graphs: An option to compare and calculate RaSE score on a sub-graph surrounding the SNP location. This option can be implemented to set vector weight(?) of distant locations to zero.

Run RaSE for a given list of position indices

Error: Reverting to path graph from sequence

Hi @fabriziocosta

With a quite high frequency for longer sequences RaSE reports the error below. For this sequence RNAplFold dotplot reports many probable basepairs.

Error in:  TTAAAGGAGAAGCCTTCAAACTCCCATAGAGCTAGATGGCTATAACTCATCTCCTCTACTCACCCTTTACTACTACCCCATTTGACCTTTTTGTAGATAACTTAGGGTTTCCATAGATATCTTTTATTAGCTCCAATGCTCCAGGTGCTTTTGTAAGTTATAATTAATTTACTCATATAGGATCCCAAAGGAAATCGCATCGCACAATGGCAGAGTTTCC
Reverting to path graph from sequence

highlight nodes that are marked * in the list (red border?)

X11 dependency?

QXcbConnection: Could not connect to display 
bash: line 1: 18755 Aborted                 HOME/Software/RaSE/1.0.0/RaSE/code/RaSE.py --draw -p 0.1 -n 15 --png -c 3 -e 2 -i UGAUGACUGACUG -w 150 -k 5 -r 0.5 -b 130

Readme.md not uptodate

it seems the "importance window" feature is not documented?!

webserver

FORNA visualization of mfe
scrollable full table output

add mfe to list for each mutation

structures plot : color only mutated node

position annotation (see #2)
mutated node in red (see #2)
mutation and wildtype letter for mutated node (see #2)
all other nodes in white

Add context length parameter of the mutation

Suppose the scenario where a sequence of length 1000 and window size is set to 200. For a mutation at position say 300, all the graph features above 500 never get impacted by the mutation. So the Rase-Score is undesirably influenced (decreased?) by extending and increasing the length of sequence beyond the window size.

The desire mode would be to let the user set a context length for the feature generation.
A less complicated option would be to fix the context length to the window size parameter.

Problems with toolz/potentially EDeN version?

I have been using the web tool, but after I checked the heatmap plots in the Jupyter notebook, I had been interested on generate heatmaps like that for a particular sequence I am interested on. I installed RaSE and Eden (the new and previous 2 versions) and I keep having this error:

AttributeErrorTraceback (most recent call last)
<ipython-input-17-306f466e9147> in <module>()
----> 1 rase_stat=rase.stability_score(0,'A')
      2 rase_stat2=rase.stability_score(12,'U')

/home/mcortesl/RaSE/code/RaSE.py in stability_score(self, pos, mutation, importance_semi_window)
    353                                pos=pos,
    354                                mutation=mutation,
--> 355                                fold_vectorize=self.fold_vectorize)
    356 
    357     def compute_stability_scores(self):

/home/mcortesl/RaSE/code/RaSE.py in stability_score(seq, pos, mutation, fold_vectorize)
    221 def stability_score(seq, pos=None, mutation=None, fold_vectorize=None):
    222     """stability."""
--> 223     vec1 = fold_vectorize([('header_placeholder', seq)])
    224     alternative = _make_variation(seq, pos, mutation)
    225     vec2 = fold_vectorize([('header_placeholder', alternative)])

/home/mcortesl/.conda/envs/rnastrtools/lib/python2.7/site-packages/toolz/functoolz.pyc in __call__(self, *args, **kwargs)
    432         ret = self.first(*args, **kwargs)
    433         for f in self.funcs:
--> 434             ret = f(ret)
    435         return ret
    436 

/home/mcortesl/.conda/envs/rnastrtools/lib/python2.7/site-packages/toolz/functoolz.pyc in __call__(self, *args, **kwargs)
    279     def __call__(self, *args, **kwargs):
    280         try:
--> 281             return self._partial(*args, **kwargs)
    282         except TypeError as exc:
    283             if self._should_curry(args, kwargs, exc):

/home/mcortesl/RaSE/code/RaSE.py in <lambda>(vec, graphs)
    155     """Curry parameters in vectorizer."""
    156     vec = Vectorizer(complexity=complexity, nbits=nbits)
--> 157     vectorize = curry(lambda vec, graphs: vec.transform(graphs))(vec)
    158 
    159     cwindow_reweight = curry(_window_reweight)(boundaries)

/home/mcortesl/.conda/envs/rnastrtools/lib/python2.7/site-packages/eden/graph.pyc in transform(self, graphs)
    302         instance_id = None
    303         feature_rows = []
--> 304         for instance_id, graph in enumerate(graphs):
    305             self._test_goodness(graph)
    306             feature_rows.append(self._transform(graph))

/home/mcortesl/.conda/envs/rnastrtools/lib/python2.7/site-packages/toolz/functoolz.pyc in __call__(self, *args, **kwargs)
    279     def __call__(self, *args, **kwargs):
    280         try:
--> 281             return self._partial(*args, **kwargs)
    282         except TypeError as exc:
    283             if self._should_curry(args, kwargs, exc):

/home/mcortesl/RaSE/code/RaSE.py in _window_reweight(boundaries, original_graph)
    135 
    136 def _window_reweight(boundaries, original_graph):
--> 137     graph = original_graph.copy()
    138     if boundaries is not None:
    139         begin, end = boundaries

AttributeError: 'generator' object has no attribute 'copy'

I have the feeling that the error is somehow related with networkx, but I already tried installing networkx 1.10 and it doesn't seem to solve the problem. Any advice on how to solve this?
Also, is there any updated installation of RaSE?

first col = rank
2.-4. col = wildtype - pos - mutation (issue #12)

store for each marked mutation an according plot

file name

RaSE_OldntPosNewnt.FILEFORMAT

maybe reasonable: swap col 1 and 2 in output

that way we get directly the mutation encoding in column style, which eases copy and paste!?

just a suggestion that we can also drop.. but for bioinformaticians it doesnt make a difference and for all others it might be useful?!

fabriziocosta / rase-rna-structural-stability-estimator Goto Github PK

rase-rna-structural-stability-estimator's People

Contributors

Watchers

Forkers

rase-rna-structural-stability-estimator's Issues

Recommend Projects

Recommend Topics

Recommend Org