soedinglab / hh-suite

Remote protein homology detection suite.
https://bmcbioinformatics.biomedcentral.com/articles/10.1186/s12859-019-3019-7
GNU General Public License v3.0
544 stars 134 forks source link

ffindex_insert_memory: Bad address #329

Closed ndmaduro closed 2 years ago

ndmaduro commented 2 years ago

Expected Behavior

No error for ffindex_insert

Current Behavior

Run the code to generate a3m custom database mpirun -np 4 /hh-suite/build/bin/hhblits_mpi -i customDB.fas -d /Uniclust30_2018_08 -oa3m customDB_a3m_wo_ss -n 2 -cpu 1 -v 0

During the process the following errors are generated:

/hh-suite/lib/ffindex/src/ffindex.c:104 ffindex_insert_memory: uc30-1808-49239845|Representativʊ: Bad address

and about 200 of:

/hh-suite/lib/ffindex/src/ffindex.c:121 ffindex_insert_memory: : Bad address

I'm more writing to clarify exactly what this error means.

From the first line "uc30-1808-49239845|Representativʊ: Bad address" this seems to indicate an error with a specific database entry? If this is the case, why is it attempting to insert c:121 over ~200 times. Does the lack of file identifier for c:121 imply that this is a blank field in the fasta file used to generate the ffindex & ffdata?

From the return to the command line entry state, I assume the database has been generated but do not want to proceed with an incomplete database or a database containing errors.

Thanks in advance for your insight.