Troubleshooting

We made STARE rather verbose, so that it tells you what it is currently doing. Also we try to catch the more obvious issues, like unavailable files, and print an appropriate error message. Nonetheless, we can’t cover all possibilities, especially since there are so many different data types coming together. We thought of some things that might cause issues and that might be worth checking, if you encounter an error. The list is not exhaustive, it is more driven by personal experience.

Index of activity columns

Make sure that all the columns you gave with the -n flags actually exist and that you start counting at 1. Otherwise it might cause an index error, C++ will call a segmentation fault.

Header in bed-files

Headers are allowed in bed-files, and also encouraged to label the activity columns, but they must be preceeded by a number sign ‘#’. Otherwise, bedtools will fail to work.

Gene annotation and gene IDs

STARE relies heavily on the gtf-file you provide. It will consider all entries where the 3rd column (1-based) equals ‘gene’ when using -i 5_tss, or ‘transcript’ when using -i all_tss. Within those entries it will take the string that follows ‘gene_id’ in the last column as identifier. The ABC-scoring step also retrieves the string after ‘gene_name’. Make sure that the format is consistent, and that gene IDs in other files are present in your gtf-file. For example, if you give a list of genes with the -u flag in the style of ‘ENSG00000223972’, but your gtf-file has ‘ENSG00000223972.4’, you won’t get any results.

The chr prefix

We rely on bedtools for intersection of regions. Bedtools requires that the naming scheme for the first column is identical, meaning that either both files have a chr-prefix, or neither of them. A mixture will cause an error. STARE will check the files in advance and should take care of this for you. But if there is an unexpected format or the first scaffolds don’t start with “chr” while the others do, this will likely fail. This affects the region file (-b), gene annotation (-a), sequence file (-g) and regions to exclude (-x). It might be worth ensuring consistency for the chromosome name in each of the files, if you suspect that to be the issue.

Is it compressed?

Almost all files are gzipped by STARE. If you decompress a file and remove the compressed version, STARE won’t find the file it needs.

Still not working?

If you still have trouble getting STARE running, open an issue on GitHub and we can try to solve it together. Also don’t hesitate to open an issue if you have suggestions, feedback or ideas for STARE.