Closed GundryLab closed 2 years ago
Thanks for picking this up -- can you try the latest on GitHub to see if it works for your data. I tried it on the test files and will take a closer look. If everything is OK I will stage for the next PyPI version.
I installed jcast using the pip command and encountered the same error. Then I downloaded the jcast folder from github and installed it using python setup.py install but the error is persisting.
I will look into this. In the meantime installing from pip then downgrading to BioPython 1.77 might be a workaround.
Hello again,
Using pip uninstall I removed the python 1.78 following which I downloaded biopython 1.77 from the source page and installed it using python setup.py install method. The installation returned no error.
Upon execution of jcast, it returned the following error:
pandas.core.computation.ops.UndefinedVariableError: name 'transcript_biotype' is not defined
Is this an error related to biopython version or is there something wrong in my command line as given below:
python -m jcast /home/jcast/diff_col/ /home/genome/gencode.v35.annotation.gtf /home/genome/GRCh38.primary_assembly.genome.fa -o /home/jcast/output/diff_col/
thanks!
The current github version should work with Biopython 1.78 but not 1.77; or at least when it did I tried a couple of datasets on my computer. If you use Biopython 1.77, you should install the PyPI Jcast release through pip, rather than install from github.
If you could send me the full error message when using the github version and Biopython 1.78 that could help me pinpoint where the problem is. Thanks!
I've attached the error as a text file and jcast log file. I'm running the Github version of jcast with Biopython 1.78 error.txt jcast_main.log
could you try using an ensembl annotation gtf to see if it works for you for the time being? I will look into what is causing the error with gencode in the next few days.
Sure, I'll try but I've used Gencode gtf and fasta for the initial alignment using STAR.
Ok - I just pushed a very quick fix to Github (Gencode uses transcript_type vs. Ensembl's transcript_biotype to denote coding transcripts in the gtf). Could you let me know if Jcast runs for your data now?
Moved to #8.
I am having a similar issue while running ductape package. Can you please suggest what I can do as a workaround?
Bio.Alphabet has been removed from Biopython
Please see: https://biopython.org/wiki/Alphabet
Because of this, JCAST does not run with the latest BioPython (1.78, released Sept. 4, 2020). The error is: "Bio.Alphabet has been removed from Biopython. In many cases, the alphabet can simply be ignored and removed from scripts. In a few cases, you may need to specify the
molecule_type
as an annotation on a SeqRecord for your script to work correctly. Please see https://biopython.org/wiki/Alphabet for more information."