Expression quantification = how is it possible that the amount of sequenced reads assigned to a specific gene/transcript

Question

[mairena.hirschberg]

Hello, I am learning about expression quantification, which according to my slides means "finding the amount of sequenced reads assigned to a specific gene/transcript". But how is it possible that more than one read is assigned to a specific gene? How are specific genes expressed more than once in reading cycle? Best wishes, Mairena

[GenoMax]

But how is it possible that more than one read is assigned to a specific gene?

Chalk it up to technology. In rare instances we prepare libraries for sequencing that do not use any amplification. In most instances there will be a shearing/amplification step in the making of libraries since we need to increase the amount of material so it becomes measurable/detectable. One will end up with overlapping library fragments in the process. This will explain multiple reads aligning to a specific gene (in addition to the "multi-mapping" that can result because of sequence similarities between domains etc). If one is worried about finding original number of molecules one is starting with then there are ways to add unique molecular identifiers (UMI) that will allow one to keep track of molecules as they go through amplification.

Answer 1

[Bastien Hervé]

How are specific genes expressed more than once in reading cycle?

I think you are mixing up gene expression biology and the technique by which one can capture this gene expression.

If a cell is expressing a gene only once during its all life, you are not gonna live for a long time...

One gene can be expressed more than a thousand times in a given cell. Then, sequencing allows you to screenshot a specific moment during the cell life to capture its transcripts. "Reads assignment" is the step where a computer is mapping the capture transcripts onto the gene it belongs to.

[mairena.hirschberg]

In expression quantification, we measure how often a cell is expressing a gene in a given moment, how is that possible? How can the same cell express a gene a few times at the same time?

[Bastien Hervé]

When you sequence a sample, you don't only capture what the cell was doing the last millisecond of its life, but you capture everything the cell has been doing before as well. Of course you cannot capture what was expressed in a cell one year before it gets sequenced because RNA (in this case) is being degradated, but you can easily sequenced RNA which has been transcribed 30 minutes before your sequencing.

Some sequencing methods are specialized into tracing what is "old" RNA to what is "nascent" RNA

You can compare that to a surveillance video footage, you cannot look at the video from the first time you pressed the record button, but you can always play the last 30 minutes.

Also, even for a screenshot at a super short timing, genes are in 2 copies in each cell and as GenoMax mentionned, for the RNA to be detectable by the machine, an amplification step is necessary.