Welcome @Danny
Yes, the tool is reporting the issue it found with the parsing.
How to troubleshoot this:
-
Review the technical details about how these tools work. In this case, the tools are from the UCSC utilities. Each tool is defined in the their website here → Genome Browser FAQ. The genePred format includes a count of the exons – since your GTF doesn’t include those features, the tool is failing to let you know that there was an unexpected input format.
Help
Converts a GTF file to a [BED12]. (Genome Browser FAQ) formatted file using UCSC tools from Jim Kent.
gtfToGenePred, followed by genePredToBed
-
Use the tool gffread to see if it can help to normalize your file into a format that the other tools can interpret better. The bed12 format expects transcripts and exons defined, along with the CDS, to produce the “block” and “thick” coordinates.
Your transcript/exons are defined in the 9th “attribute” column of the GTF but are not yet broken out into the 3rd “feature” column with their own data line.
Based on your example lines, in the final BED12 output, your thickStart/End coordinates will be equal to the blockStart/Ends.
-
While testing your files, try some of the Convert GTF to BED12 → Advanced Options to see if those can help produce that desired output. The Skip all errors option is helpful for debugging and is where I would suggest starting (after producing a normalized version of the GTF). The option for ignoring groups (aka genes) without exons will omit them from the output which isn’t what you want but you can double check my advice!
-
The message inside the dataset view in the history is just one place to see log messages. The full logs are on the job’s Details view using the i-con.
Hope this helps! 