Steps after first de novo assembly

Question

[Tong]

Hello all,

After trimming raw reads, I have tried to assemblied my NGS data by several assemblers.

As you can see in the picture, I got the different result from the different assembler. In my opinions, the assembly draft by Velvet was not so good.

The drafts by SPAdes demostrated the excellent results (N50, L50, longest contig, etc.), but the numbers of mismatches and misassemblies were not good as those of ABySS (the first and second columns).

My questions are...

1. Which one is better in your opinion?
2. Is it possible to correct the mismatches and misassemblies?
3. How can I do to improve my assembly draft?

Thank you.

Answer 1

[Istvan Albert]

Evaluating an assembly is not so straightforward.

There are a few commonly use measures, largest contig, N50 and so on.

Read up on the methodology employed by the assembly contest called Assemblathon to get a sense of the various tradeoffs that need to be accounted for

https://assemblathon.org/