Hello-
I have a test run of modifi on a metagenome running at the moment and I encountered the following:
[W::fai_get_val] Reference ctg19232 not found in FASTA file, returning empty sequence
[W::fai_get_val] Reference ctg19232 not found in FASTA file, returning empty sequence
[faidx] Failed to fetch sequence in ctg19232
[E::fai_build_core] File truncated at line 2
[faidx] Could not build fai index test/contigs/ctg19232.fa.fai
The job is still running but was curious what may have caused this one contig to be split out with only a header and no sequence:
cat test/contigs/ctg19232.fa
>ctg19232
I can confirm that this contig was not empty in the input file:
>ctg19232 length=10327 coverage=6.22 circular=no
TGACGGAATTTTTCCAGTCCGGCCGGAGCAGCCACGAGGGTCAGCGCACCATCTACCTCTGTCAGGGACAGCTTTCTCACCAGGGTTGCCTGGCCACAGTACTCTTCTGTCGGAGTCTGTGCTGCATACTGCCAGTTCATACCCCATCCCATCAGCAGAACATCTTTCGCATTCTGGAATGTAACGGCTGCATAGTTGTCGAAGCCTTCATCCAGCCATCTTGGCT......
and here is the context of the contigs immediately before and after in the input fasta file:
grep -A 3 -B 3 "ctg19232" ../contigs.fasta | grep ">"
>ctg22782 length=24670 coverage=2.28 circular=no
>ctg19232 length=10327 coverage=6.22 circular=no
>ctg22284 length=27340 coverage=4.21 circular=no
and here is how the contig appears in the .fai file:
grep -A 2 -B 2 ctg19232 ../contigs.fasta.fai
ctg18253 10830 32835805 10830 10831
ctg22782 24670 32846685 24670 24671
ctg19232 10327 32871405 10327 10328
ctg22284 27340 32881782 27340 27341
ctg21285 9656 32909172 9656 9657
On RHEL 9.4 and the command (still running) is :
modifi -r ../contigs.fasta -b hifi_reads.bc2085.bam --mge_file ../genomad/contigs_summary/contigs_plasmid_summary.tsv --threads 2 -o test
Also of note is that this was the only contig (out of the 1782 with >=5x depth) which was empty:
[2026-04-01 07:44 - main.py - INFO: 8965 contigs in the assembly.]
[2026-04-01 07:44 - main.py - INFO: 8953 contigs with depth > 0.]
[2026-04-01 07:44 - main.py - INFO: Mean depth: 4.12, Median depth: 3.24 in 8953 contigs with depth > 0.]
[2026-04-01 07:44 - main.py - INFO: 1782 contigs with depth >= 5.]
and things also look okay with the test/bams/ctg19232.bam file.
Lastly, and importantly, before I hit submit here I thought I would re-test this. In the event this was simply a filesystem glitch I re-ran this job in a new test directory (on an older OS, RH7, since that was what I had available to re-run while the first test still going ) and now this contig looks fine:
ls -l test/contigs/ctg19232
-rw-r--r-- 1 x x 10338 Apr 1 09:47 test/contigs/ctg19232.fa
-rw-r--r-- 1 x x 30 Apr 1 09:47 test/contigs/ctg19232.fa.fai
so fortunately or unfortunately I am not able to recreate the issue. And there are no other contigs for which there was an empty split fasta file. for this reason I realize this might be an quick open/close ticket but since I got this far I figure I can still submit so that there is a record of this happening at least for one user.
Hello-
I have a test run of modifi on a metagenome running at the moment and I encountered the following:
The job is still running but was curious what may have caused this one contig to be split out with only a header and no sequence:
I can confirm that this contig was not empty in the input file:
and here is the context of the contigs immediately before and after in the input fasta file:
and here is how the contig appears in the .fai file:
On RHEL 9.4 and the command (still running) is :
modifi -r ../contigs.fasta -b hifi_reads.bc2085.bam --mge_file ../genomad/contigs_summary/contigs_plasmid_summary.tsv --threads 2 -o test
Also of note is that this was the only contig (out of the 1782 with >=5x depth) which was empty:
and things also look okay with the test/bams/ctg19232.bam file.
Lastly, and importantly, before I hit submit here I thought I would re-test this. In the event this was simply a filesystem glitch I re-ran this job in a new test directory (on an older OS, RH7, since that was what I had available to re-run while the first test still going ) and now this contig looks fine:
ls -l test/contigs/ctg19232
so fortunately or unfortunately I am not able to recreate the issue. And there are no other contigs for which there was an empty split fasta file. for this reason I realize this might be an quick open/close ticket but since I got this far I figure I can still submit so that there is a record of this happening at least for one user.