Long
Note: This method has been superceded by v1.3.0 but I am retaining the below out of interest. If you want the old plassembler long method implemented from v1.3.0 onwards for some reason, you can do this with --canu_flag.
After reading this paper by Johnson et al, it seemed interesting that while most assemblers failed to recover small <10kbp> plasmids, Canu always did - albeit with multiplication. However, multplicated contigs in Canu indicate the smallest repeating sequence unit in the header, and so can be trimmed easily (see e.g. Ryan Wick's script in Trycycler).
Therefore, from v1.2.0, plassembler long will follow all the same steps as plassembler run (i.e. hybrid mode), but instead of Unicycler, it will run Canu to recover plasmids in the unmapped reads.
Of course, another big issue with long read is size selection - if your read set doesn't have many reads small enough to be a part of your <10kbp plasmids>, then nothing will work - your best bet is still to get some short read sequencing data.
But assuming they are in the read set, plassembler long should hopefully recover your small plasmids.
Results
I tested plassembler long on the 60x simulated reads I generated for benchmarking plassembler for the manuscript and can be found here. from Wick et al, C222 (Houtak et al) and Cav1217 (Mathers et al) (find more information here and here).
Overall, it seems to work pretty well - not perfect (it missed a linear and small plasmid in K variicola and also tends to assemble some non-circular chimeric contigs), but it seems anything circular is a real plasmid from the ground truth.
Other Options
You could try running Canu on the entire assembly and use Ryan Wick's script to trim the multiplicated plasmid contigs.
Ryan also thinks Canu is a pretty good assembler generally (it is perhaps more accurate than Flye). The downside is time though, I have found that it takes 5-10x more time to run than Flye on my isolates.
Results
Everything was run on my Macbook Pro M1 (2020) with 8 threads. I was too lazy to run QUAST but I would recommend polishing the output with your favourite long read polisher (Medaka) anyway. Below are the results (in terms of lengths).
By Isolate
(c) = Canu marked the plasmid as circular
| Isolate | Ground Truth | plassembler long v1.2.0 |
Time (s) | Flye (Within plassembler long ) |
|---|---|---|---|---|
| C222 | 2473 | 2473 | 1917 | Nothing - missed 2473 |
| Acinetobacter baumannii J9 | 145059; 6078 | 145058 (c); 6078 (c) | 967 | 145059; 10771 |
| CAV1217 | 181436; 70606; 44015; 9294 | 181429 (c); 70603 (c); 44015 (c); 9294 (c) | 1230 | 181433; 70605; 44015; 9294 |
| Citrobacter koseri MINF 9D | 64962; 9294 | 93661 (chimera includes 64962 plasmid); 9294 (c) | 1196 | 64961; 9294 |
| Enterobacter kobei MSB1 1B | 136482; 108411; 4665; 3715; 2370 | 136477 (c); 108402 (c); 4665 (c); 3715 (c); 2367 (c) | 1374 | 136477; 108408 - missed 3 small plasmids |
| Klebsiella oxytoca MSB1 2C | 118161; 58472; 9975; 4574 | 118159 (c); 58467 (c); 9973 (c); 4573 (c); extra 18290 (chimera) | 1646 | 118161; 58472; 9975 - missed 4574 |
| Klebsiella variicola INF345 | 250980; 243620; 31780 (linear); 5783; 3514 | 250968 (c); 243616 (c); 3514 (c) - missing linear plasmid + 5783 | 3544 | 249712; 243617; 31645 - missed 5783 + 3514 |
Overall
| Total | Missed Small Plasmids | Missassembled |
|---|---|---|
plassembler long v1.2.0 |
1 + linear plasmid | 2 (C. koseri 64962) and K oxtyoca chimera |
Flye (Within plassembler) |
6 | 1 (A. baumannii 6078) |
Summary
It's pretty good! Not perfect and I'd still recommend short reads if you can get them, but not too bad. Flye still seems the best for linear plasmids as well.
Usage: plassembler long [OPTIONS]
Plassembler with long reads only
Options:
-h, --help Show this message and exit.
-V, --version Show the version and exit.
-d, --database PATH Directory of PLSDB database. [required]
-l, --longreads PATH FASTQ file of long reads. [required]
-c, --chromosome INTEGER Approximate lower-bound chromosome length of
bacteria (in base pairs). [default: 1000000]git
-o, --outdir PATH Directory to write the output to. [default:
plassembler.output/]
-m, --min_length TEXT minimum length for filtering long reads with
chopper. [default: 500]
-q, --min_quality TEXT minimum quality q-score for filtering long reads
with chopper. [default: 9]
-t, --threads TEXT Number of threads. [default: 1]
-f, --force Force overwrites the output directory.
-p, --prefix TEXT Prefix for output files. This is not required.
[default: plassembler]
--skip_qc Skips qc (chopper and fastp).
--pacbio_model TEXT Pacbio model for Flye. Must be one of pacbio-raw,
pacbio-corr or pacbio-hifi. Use pacbio-raw for
PacBio regular CLR reads (<20 percent error),
pacbio-corr for PacBio reads that were corrected
with other methods (<3 percent error) or pacbio-
hifi for PacBio HiFi reads (<1 percent error).
-r, --raw_flag Use --nano-raw for Flye. Designed for Guppy fast
configuration reads. By default, Flye will assume
SUP or HAC reads and use --nano-hq.
--keep_chromosome If you want to keep the chromosome assembly.
--flye_directory PATH Directory containing Flye long read assembly.
Needs to contain assembly_info.txt and
assembly_info.fasta. Allows Plassembler to Skip
Flye assembly step.