Self-self blast (farm_blast bsubs itself)
farm_blast reference.fasta reference.fasta 

act reference.fasta Farm_blast.reference.fasta.reference.fasta.blast_plus.blastn.out/blast.out.gz reference.fasta &

Right-click -> ignore self matches

Look for sequence match start to end (= overlap).
Overlap of 6840bp.

Index the fasta
samtools faidx reference.fasta

reference.fasta.fai file looks like this:
unitig_0	6402387	10	60	61

Trim overlap from start of sequence
samtools faidx reference.fasta 'unitig_0':6841-6402387 > trim.fa

Find dnaA, using automated annotation
annotationfind -t sample -i 3187STDY5863006 -g dnaA -n

Use the --act flag in farm_blast to make sure all sequences are compared
farm_blast --act output.dnaA.fa trim.fa

dnaA sits at 5950817

Index the trim.fa file
samtools faidx trim.fa 
 
unitig_0|trim	6395547	15	60	61

samtools faidx trim.fa 'unitig_0|trim':5950817-6395547 > new_break.fa
samtools faidx trim.fa 'unitig_0|trim':1-5950816 >> new_break.fa

Use text editor to remove second 'contig'. Open in Artemis and re-write out all bases, to be sure it's in proper fasta format!

Find raw data with pathfind, and symlink directly to this folder
pathfind -t sample -i 3187STDY5863006 -f pacbio -l .

Re-map raw data back to this new reference, into output folder named reseq
pacbio_smrtanalysis --memory 30 --reference new_break.fa RS_Resequencing reseq *.bax.h5

Check join (at 6395547-5950817 = 444730) by looking at reseq/aligned_reads.bam over the reference sequence new_break.fa

Join looks fine, moving on to annotation (for which you use reseq/consensus.fasta)

bsub -M4000 -R "select[mem>4000] rusage[mem=4000]" 'annotate_bacteria -a consensus.fasta --sample_name PAK  --genus Pseudomonas'