Reg: Too many debris

[Harish Kothandaraman]

Hi,


I have finished the 3D-DNA pipeline, and I have gotten excellent results. The scaffolds are in the range of the chromosomes as expected. I'm sharing an screenshot as can be seen:

I do have a few questions though:


1. There are a lot of debris sequences, the stats of the input and output assemblies are here:

file	Canu_MP	Canu_MP_3D-DNA
num_seqs	3517	5026
sum_len	822873405	824737905
min_len	1011	1000
avg_len	233970.3	164094.3
max_len	18816920	80397223
Q1	22152	14647
Q2	41836	25000
Q3	124424	47000
sum_gap	175253	2039753
N50	1369959	51413987

After the scaffolding while the chromosomes are in near perfect order, I believe that the debris scaffolds are not getting places causing a lot of false-joins. I'm feeling a lot of apprehension about joining and curating 5015 scaffolds in the assembly. I believe probably there has to be a better way of integrating the same, or atleast significantly reduce the same. Would I have to rescaffold these?

2. Based on the final output from 3D-DNA there are a lot of contigs below 1Mb in size. Is it recommended to mess about with them to include them in the assembly? Because as I see it, scaffolds over 1Mb contain nearly 160Mb in the 5015 scaffolds. The parameters that can be used are the following:

• --polisher-input-size (default 1Mb)
• --splitter-input-size (default 1Mb)

3. I went through a couple of other threads where in you have suggested to change the following:

• --editor-repeat-coverage (Misjoin editor threshold repeat coverage, default 2 to 2.5 or 3)
• -r (number of iterative rounds for misjoin correction)
  
  What do you suggest do I do in this case?
  
  Any help is appreciated!
  

[Harish Kothandaraman]

Apologies,

I wanted to make a correction:

->Because as I see it, scaffolds over 1Mb contain nearly 160Mb in the 5015 scaffolds

The over should be below.