Open JuanmaMedina opened 5 years ago
@JuanmaMedina,
Depending on your use case whamG might better choice. WhamG internally classifies structural variant types, and is generally more accurate. However, if you need to classify wham calls, I'll revisit this issue.
Hello Zev,
Thanks for your fast response.
Yes, I indeed used the whamg
script, as recommended in the documentation. I edited the OP in case it was not clear.
I parsed out the resulting .vcf
to extract the SV-annotation information. However, I was trying to use the classify_WHAM_vcf.py
on the resulting raw .vcf
in case it provided me with any extra useful information. This was the reason of my issue opening. If the script actually does not provide with more information, or it is not suitable to run it after performing a whamg
run, maybe we can close this, although I think this error could persist in other users.
Cheers.
Good morning Zev,
After correctly running
WHAMG
script, I tried to perform the classification step by usingclassify_WHAM_vcf.py
. But in this step I found a couple of errors:The first is a slight modification of the source code of the python script, as the
cross_validation
module seems to be deprecated, and it is advised to import the corresponding functions frommodel_selection
instead (see https://stackoverflow.com/questions/30667525/importerror-no-module-named-sklearn-cross-validation for further details). This can be easily fixed by editing line 6 of code with:from sklearn.model_selection import cross_val_score
I have not solved the second problem, which is the main reason I am opening this issue:
Could you give me a hint here? Regarding the conda environment where I am running the script:
Thanks in advance!