nanoporetech / medaka

Sequence correction provided by ONT Research
https://nanoporetech.com
Other
391 stars 73 forks source link

Output .fasta is missing several contigs #449

Closed katholan closed 11 months ago

katholan commented 11 months ago

Issue My consensus.fasta output file from Medaka is missing almost 200 contigs (out of 1200).

The command completes successfully with no strange errors from what I can tell, but is missing significant portions of the input .fasta. The contigs it's removing range in coverage. It does copy some contigs verbatim according to the log file.

Medaka runs successfully and contains all output contigs when polishing a different assembly. The only difference between the run that outputted all contigs and the run that didn't is the input .fasta itself. The input .fastq reads, environment, settings, and commands are the same between the two assemblies. Both assemblies are from the same .fastq reads and both were assembled with Flye, just with slightly different commands.

Command

medaka_consensus -i $NANOPOREFASTQ -d $FASTA -o $OUTPUT -m r941_min_hac_g507

Log output

Checking program versions
This is medaka 1.8.0
Program    Version    Required   Pass
bcftools   1.17       1.11       True
bgzip      1.17       1.11       True
minimap2   2.26       2.11       True
samtools   1.17       1.11       True
tabix      1.17       1.11       True
Aligning basecalls to draft
Running medaka consensus
Polished assembly written to /OUTPUT/consensus.fasta, have a nice day.

Environment

katholan commented 11 months ago

I realized that the minimap step was using outdated .fai and map-ont.mmi that I forgot to delete from previous runs. Deleting these and rerunning has solved my problem.