Interpretation of results about platon HOT 1 CLOSED

Hi @franciscozorrilla,
there are many approaches to classify the replicon origin of a contig. One is the RDS method which Platon is essentially all about. However, in order to improve the classification for certain types of contigs that are hard to classify, we added several characterization steps and heuristic filters. One of these filters are mere blastn hits against RefSeq plasmid sequences. But, just because a contig doesn't align to a known plasmid sequence this does not mean that this is not a plasmid contig. For instance, this might be a contig of an unknown plasmid. This is endorsed by the detection of 2 replication proteins.

Regarding your 2nd question: In general I'd recommend the accuracy mode for any high-throughput analysis. If you can manage to take a closer look at all results in person, you could also run Platon in sensitivity or even characterization mode. The first will use relaxed RDS thresholds the latter skips all filters and fully characterizes all contigs (which might take a while).

from platon.