marbl/verkko

ONT-Only Assembly with Verkko?

Closed this issue · 1 comments

Hi @skoren ,

I am working on assembling a genome using Oxford Nanopore Technologies (ONT) R10 reads with Verkko. I have encountered a challenge with the requirement for PacBio HiFi reads (--hifi flag), which I do not have. My dataset consists of ONT R10 reads with ~100X coverage, and I would like to understand if it's possible to run Verkko with ONT-only reads.

Attempts and Observations

  • I have attempted to run Verkko using the --nano flag for the ONT reads.
  • When not supplying the --hifi flag, Verkko produces an error indicating that PacBio HiFi reads are mandatory.
  • I tried using a dummy file for the --hifi flag as a placeholder, but as expected, Verkko could not proceed with the assembly since the dummy file contained no valid reads.

Questions

  1. Is there a way to configure Verkko to run with only ONT reads, without the need for PacBio HiFi reads?
  2. If Verkko does support ONT-only assemblies, what are the correct parameters or configuration settings I should use?
  3. Are there any planned updates or workarounds that could facilitate ONT-only assemblies in Verkko? Any suggestion for another tool that can handle ONT only to produce phased assemblies? I do plan to polish the initial assembly with PE read data.

Relevant Details

  • Verkko Version: verkko bioconda 1.4.1
  • Command Used:
    verkko -d output_1000_15_decon_pcap_8735_concatenated_reads --nano decon_pcap_8735_concatenated_reads.fastq --local-cpus 24 --local-memory 600 --max-k 15 --min-ont-length 1000 --unitig-abundance 4 --hifi dummy_hifi.fastq

I would greatly appreciate any guidance you can provide. Thank you for your support and for developing Verkko.

Best,

Camilo

Verkko doesn't support ONT-only assembly unless you have duplex reads to replace the HiFi. We've been experimenting with the new Q20+ data that ONT recently released for the human genome but I expect this doesn't describe your data type. I'd look at flye or Shasta for ONT-only support.