Torsten Seemann

Yes some more sanity checking could be added, but i did assume basic QC was done at some point before Nullarbor, but yes things can still go wrong. `egrep 'warn|error'...

I have one nearly working - watch this space!

I did a lot of work on it and then never pushed it due to network problems! Nullarbor is going through some changes but there are many people like @drpowell...

Temporary hacks: 1. don't provide `--gsize 5M` as it will guess properly 2. use `--gsize=5M` instead

I will have to add a new section to the report for this. However if you use --prefill this week not be true information, because the user can patch in...

Hmmm, not sure! The code is: ``` realpath(dirname(which('mlst'))."/../db/blast/mlst.fa"); ``` In your `bash` shell can you type ``` realpath "$(dirname $(which mlst))/../db/blast/mlst.fa" ```

Both these problems are old issues that I thought were solved by multiple conda patches since. Conda often has issues with this, i think mainly to do with python version...

I am going to try and update the conda recipe and force all the newer versions of things. I think i forgot to make Kraken2 default because we use these...

That error is weird. It doesn't seem like a conflict at all. Hmmmm.

``` grep -i 'error|warn' nullarborlog.txt | less` ``` These are unusual: ``` WARNING: --max_kmer_count 10 to 14 because of high coverage for genome size 3532748 WARNING: --min_count changed from 2...