Paternal-maternal chromosome's and sequencing.

Question

[Outlander]

I am an outsider to the field,yet must know a few things about sequencing. Assume I am giving my dna to get my entire genome sequenced. So now- 1.Are the two sets of chromosomes(paternal and maternal) separated before sequencing? I understand that there is no physical connection between these two sets in the cell.Isnt it?

2.If not-I assume that sequence for both the sets of chromosome has to be provided separately or one after the other. Is it so? Or is it that only the sequence for one of them is shown in the output? When I asked this question on other forums I got stupid answers like it is completely random etc. It makes no sense to think that sequencing output would jut be a set of ATCG's without any specific order for them. Can someone please clarify. Thanks in advance.

Answer 1

[Devon Ryan]

1. With few exceptions, no, they're sequenced together.
2. The resulting data is a mix of both alleles. The "stupid answers" you apparently received previously are correct, though you're apparently missing the context within which to place them. Most common high-throughput sequencing works by randomly fragmenting the genome in a very very very large number of small (often 300-500 bases) fragments. The ends of a few tens to hundreds of millions of them are sequenced in parallel. There's no particular order or preference to this, which leads to the "completely random" part that you were told before.

[GouthamAtla]

"With few exceptions" . Is there any way to sequence separately ? Interested to know.

[Devon Ryan]

Historically speaking, people would make human-mouse hybrid cell lines with just a single human chromosome. That was still common practice when I was a grad. student (so, still within the last 10 years). These days it's probably easier to just pay a bit more to use the 10x Genomic library prep. That's not as good, but it'll make doing the phasing a LOT easier.

[Outlander]

https://www.technologyreview.com/s/422149/sequencing-a-single-chromosome/ Now that is something which I came across a few days ago. No idea whether it helps.

[breckuh]

Could you just sequence 1 parent and then use the reads in the child that match reads in the parent to separate the 2?

[Outlander]

Thanks Devon. So I am to understand that the sliced bits of dna from both the sets of chromosome are put together without following any specific rule.Isnt it? Assuming for simplicity that the dna is sliced region wise-in the output 1p22 sequence may be followed by 7q12 sequence. Is it so? Now since actually slicing doesn't happen region wise and that a gene region may be cut into pieces wouldn't the entire process be meaningless if they are not put together in order. Somehow I find this a bit hard to grasp. It would be of great help if you can take some time to clarify. ...................... Assuming slicing happens chromosome wise the idea which I have now in mind is- Chrom1(paternal)-chrom2(maternal)-chrom3(paternal/maternal) etc Chrom1(maternal)-chrom2(paternal) etc. Here I see there should be two output sequences-which isn't the case. Or is it- Chrom1-Chrom1-Chrom2-Chrom2 etc or ATCGTACG etc with alternate base from paternal and maternal set. May be it is all stupid.But I have spend past 3-4 days trying to find some help on this.Awaiting reply.

[Devon Ryan]

The small bits ("reads") aren't put together at all. The sequencing machine gives you one or two files with 4 lines per read ("fastq format"), with the ordering in the file being however the reads happened to randomly be arranged when they annealed to it. That's it. You can then assemble those short reads into a genome, or, since you're interested in sequencing human DNA, map it to the reference genome and then "call variants" (i.e., make a list of the differences between what you sequenced and what the reference sequence is).

[Outlander]

Thanks Devon. Now I got it.Maybe others didn't mention it as it is such a basic thing in their eyes.