Omni-C scaffolding

[ibrc.ak...@gmail.com]

Dear 3D-DNA developers,

First of all, thank you for the excellent software.

I am trying to scaffold a plant genome (570Mb) using Omni-C reads. 

I tried using the default settings with an option "--editor-repeat-coverage 5", but it doesn't seem to be working very well.

Is my Omni-C library not working? 

Can I change the parameters of 3D-DNA to get a promising scaffold?

It would be nice of you to share your experiences and opinions.

I attached inter.txt, 0.hic and final.hic files.

Thanks for your help,

Akira

[Olga Dudchenko]

Hello Akira,

Indeed you seem to have a lot of variation in coverage across your library. Modifying the editor repeat coverage is indeed the standard way to deal with that. I would recommend examining the source of coverage variation: is it the undercollapsed heterozygosity, is it heterochromatin etc. (The library itself appears to be fine: my guess the source of the problem lies in the draft, so might be useful to investigate.) Lookint at your rawchrom map it seems that the scaffolding is doing more or less ok, it is the chromosome splitting bit that you want to potentially modify. 

The parameters related to identifying boundaries of chromosomes are as follows:

**splitter**

--splitter-input-size splitter_input_size

Splitter input size threshold. Scaffolds smaller than polisher_input_size are going to be placed into unresolved (Default: 1000000).

--splitter-coarse-resolution splitter_coarse_resolution

Splitter coarse matrix resolution, should be one of the following: 2500000, 1000000, 500000, 250000, 100000, 50000, 25000, 10000, 5000, 1000 (Default: 25000).

--splitter-coarse-region splitter_coarse_region

Splitter triangular motif region size (Default: 3000000).

--splitter-coarse-stringency splitter_coarse_stringency

Splitter stringency parameter (Default: 55).

--splitter-saturation-centile splitter_saturation_centile

Splitter saturation parameter (Default: 5).

--splitter-fine-resolution splitter_fine_resiolution

Splitter fine matrix resolution, should be one of the following: 2500000, 1000000, 500000, 250000, 100000, 50000, 25000, 10000, 5000, 1000 (Default: 1000).

I cannot judge from the static image what is going on, but you might want to try to split at a lower coarse resolution. Note that you don't have to rerun the whole pipeline and can start from splitting using the -s|--stage option.

Hope this helps and good luck,

Olga