louadi / nease Goto Github PK

Hi
I was wondering whether you would be able to add Mouse database into this tool or is there a way I could manually do it?

Will be much appreciated.

Thanks

Dear developers,
Thanks a lot for your creative work for AS analysis. I have been searching around for such tool and had finished the AS findings by MAJIQ and rMATS. When I studied the tutorial for MAJIQ input, I noticed that the colnames of output of MAJIQ be like:
Gene ID LSV ID LSV Type E(dPSI) per LSV junction P(|dPSI|>=0.20) per LSV junction P(|dPSI|<=0.05) per LSV junction RP E(PSI) MP E(PSI) A5SS A3SS ES Num. Junctions Num. Exons Junctions coords IR
And the colnames of newest version of MAJIQ is :
gene_id lsv_id lsv_type mean_dpsi_per_lsv_junction probability_changing probability_non_changing A_mean_psi C_mean_psi num_junctions num_exons junctions_coords ir_coords
The difference made it hard to go on and I found that the main distinction is that new version deleted the "A5SS A3SS ES" columns. In the process.py of NEASE, if data['ES'].dtype=='bool':data=data[ data['ES']==True] else:data=data[ data['ES']=='True'] seems like just extracting the exon skipping events for downstream analysis, which is hard to figure out the type of complex LSV in new version of output. Could I rename the columns of my files and create a new column "ES" with all true values ?
Thanks for any advice.

Dear Louadi,

For our project, we want to compare the number of splicing events obtained by MAJIQ with RMARTS. Then, we want to see the differences in results in NEASE between these two tools but there is a format issue on the RMARTS table. We get the following error: "process failed....Try to use the Standard input". We have looked at the code and we have the required columns, I attach the rmarts data. Can you tell me what exactly the expected format is?
SE.MATS.JCEC.txt
SE.MATS.JC.txt

Thank you for your help,

Marie

Hello again,

I am wondering if there is a way to extract which splicing events are predicted to disrupt the ORF/lead to NMD. I have looked through the source code and there doesn't seem to be an easy way to do so.

Best, Alex.

Hello,

I have been using NEASE (it seems great so far) and had a quick question. I am wondering whether it is possible to use additional ontologies such as the human phenotype ontology with NEASE. If yes, any suggestions on how I could do that?

Best, Alex.

When running NEASE with newest version of pandas, get error due to the fact that append() function is no longer present in the new pandas. This occurs after some results have already been printed out. It would be nice to either explicitly mention that a lower pandas version is required or update code for new pandas. Unless there is a fix that I am missing. Thanks!

Traceback
{
"name": "AttributeError",
"message": "'DataFrame' object has no attribute 'append'",
"stack": "---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
~\AppData\Local\Temp\ipykernel_33804\2553761319.py in ?()
----> 1 events = nease.run(data_for_nease, organism = 'Human')

\Lib\site-packages\nease\nease.py in ?(self, data, organism, input_type, p_value_cutoff, remove_non_in_frame, only_divisible_by_3, min_delta, Majiq_confidence, only_DDIs)
227
228 # Identify binding of affected domains = Edges in the PPI
229 # get DMI adn DDI for spliced domains
230
--> 231 self.interacting_domains=affected_edges(self,Join,self.only_DDIs)
232
233
234

\Lib\site-packages\nease\functions.py in ?(nease_data, Join, only_DDIs)
117
118
119
120
--> 121 interacting_domains=interacting_domains[['Gene name','NCBI gene ID','Identifier','dPSI','Affected binding (NCBI)']].append(elm_affected, ignore_index=True)
122
123
124

\Lib\site-packages\pandas\core\generic.py in ?(self, name)
5985 and name not in self._accessors
5986 and self._info_axis._can_hold_identifiers_and_holds_name(name)
5987 ):
5988 return self[name]
-> 5989 return object.getattribute(self, name)

AttributeError: 'DataFrame' object has no attribute 'append'"
}

Hello,

I have been using NEASE and had a quick question. I am wondering whether it is possible to use additional ontologies such as the MSigdb Hallmark. If yes, do some suggestions on how I could do that?

Meanwhile, I noticed that currently NEASE can only handle SE type events for rmats input, right?