Hello,
I have assembled a bacteria (4.5Mb) de novo (illumina; Paired-End; N50: 103789; L50: 12), and I want to optimize the assembly.
What program can you suggest to me to reduce the number of contigs, concatenating (or overlapping) them, without using reference?
I prefer not to use reference genomes at this stage because I have noticed that some contigs with resistance genes are not included (depending on which reference genome you choose).
Thanks,
There are many genome assembly tools / assemblers available which can perform de novo assembly. You can choose according to your organism. It is fairly easy and fast process.
But as you said you already prepared assembly, it should have merged overlap regions. How did you do it?
Hi Abhijeet Singh and Mathew A Beale, I have 3 Illumina libraries (paired-end), all the libraries together I have assembled with:
SPADES (--careful -k 21,31,41,51,61,71,81,91),
SPADES (--plasmid --careful -k 21 , 31,41,51,61,71,81,91),
VelvetOptimiser (--s 21 --e 91 --x 10) and
Velvet (kmer 91; -cov_cutoff auto -ins_length 250 -exp_cov auto -scaffolding no) , but SPADES gave me the best metrics.
Among the contigs that SPADES generated, I was able to semi-manually circularize (a few contigs) a possible plasmid (9Kb + -) selecting the contigs with greater depth. So I have doubts if SPADES really manages to merge (merge, concatenating or overlapping) the ends of the contigs or maybe it is not very flexible. Is it possible to concatenate or overlap all the contigs obtained from the various assemblers? perhaps this way I recover longer segments of the bacteria.
Mathew A Beale thanks for the information I will review it.
- I think you should use Unicycler. It uses spades for assembly and perform k-mer optimization automatically. It also uses Pilon, which automatically improve draft assemblies and thus provides an improved representation of the genome. It can be visualized in Bandage to interact with the assembly graphs.
- Badges
- Science topic
- Similar topics
- Filing