EM not converging with default database

[aaguinal]

I am using the default minSeq+H database. I have run the following commands to classify a FASTQ file:

metamaps mapDirectly --all -r /metamaps_db/default/databases/miniSeq+H -q input.fastq -o /output_dir
metamaps classify --mappings /output_dir --DB /metamaps_db/default/databases/miniSeq+H

However, the standard output shows the following endlessly:

EM round 29017654                                                                              

        Log likelihood: 0
        Improvement: 0
        Relative   : -nan
EM round 29017655

        Log likelihood: 0
        Improvement: 0
        Relative   : -nan
EM round 29017656

        Log likelihood: 0
        Improvement: 0
        Relative   : -nan
EM round 29017657

        Log likelihood: 0
        Improvement: 0
        Relative   : -nan

It seems that the EM algorithm is not converging and/or updating.

I running this using the Metamaps BioContainer using the Docker imagequay.io/biocontainers/metamaps:0.1.98102e9--h176a8bc_0.

[AlexanderDilthey]

Hi @aaguinal,

I have never come across this issue before - could you post the contents of your /output_dir.meta file?

Best,

Alex