Forward and reverse reads in NGS

[tahamasoodi]

Hi,
What does forward and reverse reads mean in NGS, is it like 5' to 3' and vice versa? If it is like that then the 2 reads should be complementary to each which is not the case.

[HESmith]

Forward and reverse are relative to the reference genome sequence.

[krobison]

Depending on fragment size, the forward and reverse reads may or may not overlap.

(if we ever do a FAQ, this needs to go in there!)

[tahamasoodi]

For a fragment size of 500bp, it will have 5 forward reads and 5 reverse reads of 100bp each. But each pair of reads (forward and reverse) should have the same complementary sequence.

[GenoMax]

For a fragment size of 500bp, it will have 5 forward reads and 5 reverse reads of 100bp each. But each pair of reads (forward and reverse) should have the same complementary sequence.

If you could exactly split your 500 bp fragment into 5 contiguous pieces and then sequence each of them from both ends for 100 cycles then it will be like that.

In fact for sequencing/re-sequencing you expect to get a population of fragments so that you can cover the entire fragment/region/chromosome/genome you are trying to sequence.

[tahamasoodi]

Thanks! For a single fragment of 500 bp, sequenced for 100 cycles, what does forward and reverse strand mean? Will they cover only the ends of this fragment?

[GenoMax]

Thanks! For a single fragment of 500 bp, sequenced for 100 cycles, what does forward and reverse strand mean? Will they cover only the ends of this fragment?

The forward and reverse reads will sample 100 bp from each end of the two strands of this fragment (if 100 cycles each were run on both ends, otherwise you will have 50 bp from each end if you ran a total of 100 cycles).

[tahamasoodi]

How about the middle 300 bp segment?

[HESmith]

Surely you are capable of figuring out the answer! Draw a picture if necessary.

[krobison]

For a fragment size of 500bp, it will have 5 forward reads and 5 reverse reads of 100bp each. But each pair of reads (forward and reverse) should have the same complementary sequence.

I think you may need to specify your problem a bit better. Start with you input DNA and walk through all the shearing, ligation & so forth steps.

For example, if you have a 500bp amplicon with a fusion design (sequencing adapters in the primers), then you will always read the same 100bp end region from the forward primer and the other end's 100 bp with the reverse primer.

If same situation, but you ligated oligos instead of fusion, then each primer will read a 50:50 mix of each end.

If you had a 500 bp amplicon and sheared it, then ligated adapters, you will get a population of reads which in theory can be assembled into the original fragment. If your fragment size after shearing is <200bp, then some overlap is expected

If you have sheared genomic DNA into 500 bp fragments, you will read 100 bp from each end of each fragment, but the fragments should be random sample of genome (and randomly oriented).

[tahamasoodi]

Actually, I've whole genome data generated by Illumina Hiseq 2000 with read length of 100 bp. Now while viewing the data in IGV, I got confused when I choose the option "Go to Mate". I was thinking the mate of a read will have the complementary sequence but it is not the case and some reads have mates present in different chromosomes.

[swbarnes2]

Actually, I've whole genome data generated by Illumina Hiseq 2000 with read length of 100 bp. Now while viewing the data in IGV, I got confused when I choose the option "Go to Mate". I was thinking the mate of a read will have the complementary sequence but it is not the case and some reads have mates present in different chromosomes.

Honestly, you are going to have a very hard time working in this field unless you have at least a Freshman college student's understanding of molecular biology. This is not the place to acquire that knowledge.

So no, we do not expect paired read to be the same sequence just in opposite directions. They should be about a couple hundred bases apart, pointing at each other.

And yes, in theory, you should not have a pair of reads that map to different chromosomes. But in the real world, reference sequences are not perfect, genome are not 100% unique at every point, and weird stuff just happens when you are collecting millions and millions of data points.