search

Xinglab / espresso

Other
48 stars 4 forks source link

ESPRESSO_S dies #8

Closed paulimer closed 1 year ago

paulimer commented 1 year ago

Hi all, I was testing ESPRESSO on a single sam file, but the ESPRESSO_S command seems to die at the same point of the .pl file. The <IN> line number varies by run. Thread 1 terminated abnormally: Died at /home/etheimer/ESPRESSO/src/ESPRESSO_S.pl line 865, <IN> line 88957.

The command run is the default one : 

perl ~/ESPRESSO/src/ESPRESSO_S.pl -L ./samples_espresso.tsv -F ~/ref_proj/Homo_sapiens.GRCh38.dna_rm.primary_assembly.fa.gz -A ~/ref_proj/Homo_sapiens.GRCh38.108.gtf -O ./espresso_results

Thanks in advance.

EricKutschera commented 1 year ago

It looks like this is the line for the error: https://github.com/Xinglab/espresso/blob/v1.3.0-beta/src/ESPRESSO_S.pl#L865

I think that would happen if the chromosome found in the .sam file was not in the provided .fa file

Your command uses Homo_sapiens.GRCh38.dna_rm.primary_assembly.fa.gz. ESPRESSO expects the file to be unzipped and doesn't check for a .gz file. Since it didn't read the sequences from the .fa file, ESPRESSO fails to match the chromosome names when reading the .sam file

paulimer commented 1 year ago

It was indeed the issue! I don't know why I assumed gzipped fasta was okay. Thanks for the quick answer!