Your biological sample contains a number of distinct genomes and you want to know the number of distinct genomes, their exact identities, and their relative abundances.
CluCon is analysis code that clusters reads and estimates consensus from PacBio sequencing data from mixed populations to answer these questions. It is designed for mixed genomes (or genome regions) where long PacBio reads can cover the entire genome.
Here is an example data analysis run that examines a mixture of near-full length HIV genomes (9kb long): README_HIV-three-clones.html
PacBio can sequence entire HIV genomes from single molecules as single, continuous 9kb+ reads. Given a sample containing an unknown number of HIV genomes, one PacBio sequencing chip in three hours, and the CluCon software, we were able to determine that the sample contained three full-length HIV species in a (60%/20%/20%) mixture and got the exact genomic identity of all three species.
The calling sequence is simple taking the sequence data and a generic reference to seed the process:
ConsensusClusterSubset.py \
--nproc=1 \
--runDir=bound-clucon-dna622 \
--fasta=raw-niaid-dna622.fasta \
--ref=hiv_hxb2_whole_genome-covered.fasta \
--spanThreshold=99.0% \
--entropyThreshold=1.0 \
--basfofn=dna622.baxh5.fofn