ratschlab
Several warnings using the test-data
Dear Yi Zhong (@yizhong),
After trying to run your tool and checked the opened/closed issues on GitHub while I was trying to solve why I was getting some errors/warnings, I decided to try to run the tool with the test-data (apart from the toy one) to see if something related with the tool or my data was causing the problems.
I am aware that a similar issue related to warnings have been posted (https://github.com/ratschlab/RiboDiff/issues/4), but since they were talking about own samples and not test-data, I decided to post a new issue.
I used this test-data and although I am supposed to get the same log file as the one that you uploaded, I am getting the same warnings as they have mentioned in this issue.
In that particular issue, it is mentioned that they were working with own data (and therefore, it could be something related to the format, number of samples, etc.) but since I am getting the same errors working with the test-data, I am wondering what could be. I installed the tool through github (not conda) and I loaded the python 2 module specifically before launching the tool (module load python/2.7.18)
These are the warnings that I received running the following command:
TE.py -e ./test-data/experiment_design.csv -c ./test-data/read_count_table.tab -o ./test-data/Results_TestsData/results.tab -d 0 -r 1 -p 1
(note that TE.py is in my PATH and . is the directory where I saved everything related with Ribodiff)
*************************
Read input files: Done.
1693 Gene(s) to be tested.
*************************
Library size:
['RFcontrolRep1' 'RFcontrolRep2' 'RFtreatedRep1' 'RFtreatedRep2']
[1.127 0.801 1.096 0.955]
['RNAcontrolRep1' 'RNAcontrolRep2' 'RNAcontrolRep3' 'RNAtreatedRep1'
'RNAtreatedRep2' 'RNAtreatedRep3']
[1.019 1.311 0.972 0.769 1.091 0.919]
*************************
Start to estimate raw dispersions.
/mnt/netapp1/Optcesga_FT2_RHEL7/2020/software/Compiler/gcccore/system/python/2.7.18/lib/python2.7/site-packages/scipy/optimize/_minimize.py:761: RuntimeWarning: Method 'bounded' does not support relative tolerance in x; defaulting to absolute tolerance.
"defaulting to absolute tolerance.", RuntimeWarning)
1693 genes finished...
*************************
/home/usc/gr/eer/.local/lib/python2.7/site-packages/RiboDiff-0.2.1-py2.7.egg/ribodiff/fitdisp.py:35: RuntimeWarning: invalid value encountered in greater
idx = np.logical_and(dispRaw > lowerBound, dispRaw < upperBound).nonzero()[0]
/home/usc/gr/eer/.local/lib/python2.7/site-packages/RiboDiff-0.2.1-py2.7.egg/ribodiff/fitdisp.py:35: RuntimeWarning: invalid value encountered in less
idx = np.logical_and(dispRaw > lowerBound, dispRaw < upperBound).nonzero()[0]
/mnt/netapp1/Optcesga_FT2_RHEL7/2020/software/Compiler/gcccore/system/python/2.7.18/lib/python2.7/site-packages/statsmodels/genmod/generalized_linear_model.py:273: DomainWarning: The identity link function does not respect the domain of the Gamma family.
DomainWarning)
/home/usc/gr/eer/.local/lib/python2.7/site-packages/RiboDiff-0.2.1-py2.7.egg/ribodiff/fitdisp.py:49: RuntimeWarning: invalid value encountered in less
if np.nonzero(dispFitted < 0)[0].size > 0:
Fit dispersion: Done.
*************************
Start to estimate adjusted dispersions.
1693 genes finished...
Estimate dispersion: Done.
*************************
Start the statistical test.
1693 genes finished...
Warning: Failed to do test: 0 genes. P value set to 'nan'.
Statistical test: Done.
*************************
Calculate TE and fold change: Done.
*************************
Write output file: Done.
*************************
Save data: Done.
*************************
Make plots: Done.
*************************
Why do I get those warnings with the test data?
I do have to mention the last warning that also appears almost at the end cause I have seen it a lot working with my own data: Warning: Failed to do test: 0 genes. P value set to 'nan'.
Note that I am using Python 2.7.18 and:
- numpy 1.16.6.
- scipy 1.2.3
- matplotlib 2.2.5
- statsmodel 0.10.2
Any help/comment will be really appreciated.
Thanks very much in advance!
Regards
