how to specify -n for auto-allohexaploid

[pashapu43]

Hi

I am trying to assemble a hexaploid plant genome (AAAABB, 2n=6x=42) with haploid genome size (n) estimated to be 4.25Gb by flow cytometry. Though unclear, it is believed to have resulted from the hybridization of an autotetraploid (2n=4x=28) and a diploid (2n=2x=14). How do I specify the -n option in this case?

Could you explain a bit more about using Ultra-long data? Can I input ONT reads when running the cphasing pipeline?

Regards
Akhil

[wangyibin]

Hi, sorry for the delay in my response.

Your case is quite complex, and we haven't encountered a similar example before. However, we suggest the following approach:

1. Initial grouping: Set -n 0:0 to let the program perform automatic grouping.

2. Check clustering pattern: Use the heatmap to determine whether subgenomes A and B are clustered together or separated.

   • If A and B subgenomes cluster together, treat them as homologous and use -n 7:6 to specify the expected grouping.
   • If the automatic grouping results in approximately 14 groups, and the second-level grouping includes 2–4 (or more) subgroups per group, you can refine this step under the 3.hyperpartition/ directory.

3. Manual hyperpartition refinement:
   Modify the hyperpartition.cmd.sh script to manually define target groups by specifying the -n and -f parameters:

   -n 14:second.number.list -f first.cluster.txt
   

   The file passed to -f should be a tab-delimited file with two columns:

   • The 1-based index of clusters in first.cluster.txt
   • The corresponding group number you want to assign

The ultra-long read mode is still experimental and currently not recommended for general use.

Best wishes,
Yibin