help regarding abyss assembly

Question

[praasu]

Hi,

I have paired DNA-seq data of some protista. I tried to do assembly using the abyss 2.1.1 after trimming using Trimmomatic. Trimmomatic preprocessed reads into Paired end reads (Forward and Reverse) and unpaired reads from each file that I have considered as single end reads for abyss assembly.

Used command for the abyss

for k in {150..250..10};
do 
    mkdir k$k;
    cd k$k; 
    abyss-pe k=$k name=1604 pe='../protista_forward_paired.fastq ../protista_reverse_paired.fastq'  se='../protista_forward_unpaired.fastq ../protista_reverse_unpaired.fastq'; 
    cd ..;
done

I got three contig files like protista-1.fa, protista-2.fa, protista-3.fa.

complete list of output :

coverage.hist,
protista-bubbles.fa,
protista-1.fa,
protista-1.dot,
protista-1.path,
protista-2.dot1,
protista-2.fa,
protista-2.dot,
protista-3.dot,
protista-2.path,
protista-3.fa,
protista-indel.fa,
protista-unitigs.fa ,
protista-3.fa

I am not sure which config file I should use for further analysis including scaffolding. Could someone please help me with this.

Thank you very much in advance.

Answer 1

[colindaven]

You can run some stats on each of the .fa FASTA files there to check which one looks the best.

I recommend

stats.sh

from the bbmap package, available in bioconda. It will give you lots of stats on the assembly #bp and #contigs, + N50 etc.

[praasu]

Hi,

Thank you very much for your reply. I'll check the assembly statistics for each K mer output. I will select the best k-mer based on the stat. My question is that I have used both paired end and unpaired data for the abyss assembly. I got three contig files (protista-1.fa, protista-2.fa, protista-3.fa) However I was expecting single file. So I want to know, where I will use those file for scafolding or unitigs file (protista-unitigs.fa).

[praasu]

Meanwhile, I just notice that contig file (protista-3.fa) and unitigs file (protista-unitigs.fa) are same. They have signficantly higher N50, N80 as well as other parameters. Thank you very much.