What are Tiles in Sequencing

Question

[a.krassnig]

Hello, this might be a rather simple question but I did not quite understand what I found on the Internet about it.

Looking at next generation sequencing I came across the term "Tile". What exactly are those tiles and why does it seem they always have a length of about 250-300bps?

The only thing I can imagine, ist that for example I want so sequence something with the length of 1800bps this would be way to long for the typical Illumina sequencer. So I "split" it up in to 6 tiles with the length of 300?

It would help me a lot, if someone could explain to me what exactly those tiles are in next generation sequencing, since I can not figure it out.

Thank you very much.

Answer 1

[GenoMax]

See my answer (and other comments) in this thread for an explanation of tiles: Significance of Tile in Sequencing?

[a.krassnig]

Thank you for the answer. So if I got this right, the tiles are just physical places on the flow cell? But this still does not quite answer my question why it seems that sequences are split up in to tiles.

Or is it like this:

I have sequences of 600bps. I split those up in to two sequences of 300bps. -> Each of these sequence-parts i then "place" on the flow cell on a designated Tile. So in the end I would have: Tile 1 first 300 bases of the sequences Tile 2 second 300 bases of the sequences

Is this the idea of tiles?

Thank you in advance:)

[GenoMax]

Tiles are simply large areas on a flowcell that software follows. Each tile contains a certain number of clusters (which are formed by clonal amplification of an individual library fragment). These clusters are followed over time (using their X,Y coordinates in that tile) by the sequencer software that keeps track of them (think of stars in a starry night like astronomers have done for a long while). Each cluster produces one base call for every cycle/round of "sequencing by synthesis". Number of sequencing cycles is what decides the length of the sequence. Tiles have no relation to that length.

Sequences are not split up by tiles. Sequence of a cluster in a particular tile will be split in respective R1/R2 files (but it will always be in the same tile in the two files) depending on whether you are doing single-end or paired-end sequencing.

[a.krassnig]

Ah, I think I am starting to understand it.

But as an example: in this experiment the Pseudomonas aeruginosa wt-sequence is divided in to Tiles. And i checked, if you stack the sequences in this tiles you get the wt-sequence (except for some variants of course) This is why I was thinking, it is split up by sequences...

I think the main question I still have is why the sequences in one tile are mostly quite the same?

[GenoMax]

I don't know why they used the word tile in this submission. I don't think the tile here is representative of what we have been talking about. Since I don't see original Illumina format headers in the data we can't verify anything about actual tiles. I am not sure what the residue numbers (e.g. tile 2 (residues 85-170)) are referring to either since the reads all appear to be 300 bp.

I could not link the SRA ID's to a publication but if you know of an associated publication then I suggest you check that for they have exactly done with this data..

Answer 2

[swbarnes2]

As far as I can tell, MiSeqs only have two tiles. That particular submission seems to be using the term "tile" to mean something other than "section of an Illumina flowcell"