FastQC is a super easy free tool to look at the quality of your reads.

http://www.bioinformatics.babraham.ac.uk/projects/fastqc/

I used pandaseq for this assembly. Actually, I want to know whether this assembly is efficient or not.

There are several ways to check the quality of an assembly. When you are comparing two assemblies the most common parameters to use are: N50, L50 and assembly size. Nevertheless these parameters doesn't give you an idea about chimeras and missassemblies. To evaluate that you can use some tools such as:

+ REAPR (https://www.sanger.ac.uk/resources/software/reapr/)

+ QUAST (http://bioinf.spbau.ru/quast)

Check the article from El-Metwally, M. Ouda and Helmy, 2014 (SpringerBriefs in Systems Biology Volume 7, 2014, pp 95-101)

FastQC is a super easy free tool to look at the quality of your reads.

http://www.bioinformatics.babraham.ac.uk/projects/fastqc/

I used fastqc, kmer content and per base GC content were failed

That alone doesn't mean much, since it depends on what you are sequencing. The per base seq quality is the most useful measurement as it gives you the quality scores at each position of your reads.

I used fastqc in the past and I always see there was something wrong in the K-mer and GC content, I don't really know why. Indeed, even though the quality of the reads were good, in the first few base pairs (7-10 bp) of the reads there was always some strange fluctuation if the functions. These fluctuation caused the parameter fail. Could it be due to the algorithm, not particularly efficient to calculate these trends?

Let's see the GC content (I don't know how it works, I'm just guessing). It seems that the algorithm starts from the first base of every read and makes a cumulated average of the GC content as it moves along the sequence. Of course the function fluctuate in the first part. The more is small a sequence, the greater is the weight of every nucleotide...

Does it make sense? What do you think of?

I'm sorry to insert another question in this topic but I think it could be interesting for Jingjing Peng too to know how fastqc works.

If the problem is in the first 10 bp of your global reads, just trim them and re-assess the quality. This usually fixes the problem. But keep in-mind that this can affect the quality of the subsequent mapping.

Riccardo Rinaldi, it is exactly similar to what you said. The first base pairs of the reads is always strange. I agree with Alexandre Marand. We can trim it. Did you test the change of the function due to this strange first base pairs?

No, I did not because they told me they were good.

Indeed, I checked the alignment I carried out with a reference genome and it was fine. I saw no errors on the first few bases (using Tablet), and the alignment values were good.

I see no reasons to trim them out and lose information, particularly because it seems the algorithm is just inefficient.

You can use prinseqlite

http://prinseq.sourceforge.net/

I agree. It is very useful.