Problem about：Multi-sample vs. Merged Single-sample Assemblies

[try-hunter]

Hello, I have a question. If I have two paired-end samples from the same cohort and run the following command:

spades.py --pe1-1 SampleA_R1.fq --pe1-2 SampleA_R2.fq \
          --pe2-1 SampleB_R1.fq --pe2-2 SampleB_R2.fq

How will the resulting scaffolds differ from those obtained by assembling each paired-end dataset separately and then concatenating the scaffolds together?

In other words, does adding multiple samples simply concatenate the scaffolds from each sample, or does it affect the final assembly result?

[asl]

What do you mean as "concatenating"? If these are same species, then just blindly concatenating scaffolds you would duplicate everything.

[try-hunter]

If I have two metagenomic samples, A and B, with paired-end sequencing:
SampleA_R1.fq SampleA_R2.fq SampleB_R1.fq SampleB_R2.fq

Using:

--meta --pe1-1 SampleA_R1.fq --pe1-2 SampleA_R2.fq \  
      --pe2-1 SampleB_R1.fq --pe2-2 SampleB_R2.fq \

produces scaffoldA+B.

Then, assembling them separately:

--meta -1 SampleA_R1.fq -2 SampleA_R2.fq  
--meta -1 SampleB_R1.fq -2 SampleB_R2.fq  

produces scaffoldA and scaffoldB.

Is scaffoldA+B equivalent to scaffoldA + scaffoldB? Or is scaffoldA+B a meaningless results.

[asl]

Well, it all depends what do you consider as "meaningless". Consider you combine input DNA from these two samples and then sequence them together. Are the obtained results meaningless?