Hi
I'd like know what is the difference between the terms reads and tags, aligment tags in relation to NGS.
Thanks!
•
written
by
JorgeGonzalez •
I have seen this term used quite a bit with ChIP-Seq. IIRC it is generally synonymous with 'fragment'.
EDIT: for an example where 'tag' is used in ChIP-Seq see the original MACS paper.
•
written
by
Chris Fields
Therefore, tags is referred to aligned fragments?
•
written
by
JorgeGonzalez •
Yes if you are using SE reads. Re: PE reads, I recall reading somewhere that a tag would refer to the fragment that is immuno-precipitated and thus would be essentially counting pairs of properly-mapped reads. In other words, if you had 20 properly-paired reads (10 R1 and 10 R2), you would have 10 fragments or tags.
Maybe it was defined in one of the peak-calling papers or maybe ENCODE methods, not sure, but this is logically similar to RPKM vs FPKM (RNA-Seq amongst others) so makes sense.
EDIT: this link has an answer from Sean that explains its origins a little more. I personally don't like nor use the term.
•
written
by
Chris Fields
Here an answer for tags in the context of small RNA-Seq:
When doing small RNA-Seq to sequence short mature microRNAs, one will realize that the reads are not unique after adapter clipping. Due to the Drosha/Dicer cuts, one ends up with thousands of copies of reads having exactly the same sequence. These sequences can be used to quantify the microRNA. In contrast to mRNA-Seq, one would never try to get rid of these multiple copies of reads, since they hold information in small RNA-Seq experiments.
Now to your question: When doing small RNA-Seq, these multiple exact copies of a read are merged to one tag. One tag consists of thousands of reads. This way, one can decrease the problem. Instead of mapping one sequence several thousand times to the same positions, only one tag, representing thousands of reads, is mapped.
Log in to answer this question.
Are you thinking of ESTs? Expressed Sequence Tags? Those are usually reads that represent a portion of the transcriptome.