Hi, when proofreading for a colleague I just stumbled upon an unreferenced claim that we have thousands to millions of sequence copies in each cluster using Illumina technology.
Wikipedias Illumina dye sequencing can't decide between "about a thousand" and "hundreds" referencing just one article from 2015 using MiSeq which is behind a paywall.
I found another article from 2008 claiming "approximatly 1000".
But I can't imagine this numbers are still accurate for current technology including two-colour chemistry. If the amplification is not restricted too much by reagents or binding sites on the flowcell, 35 cycles of cluster generation for v4 Chemistry should generate a lot more.
Does anybody know the numbers or have a reference for current technology systems/flowcells/chemistry?
I don't know the answer, but you can't grow exponentially when bounded on a 2-dimensional plane, only quadratically with time.
Hi Brian, I don't know the details on a flowcell, but never thought of the process as just "growing one step 2-dimensinal" like "occupy the closest field on a chess-board like the pawn". I always assumed sequences to be long enough to jump a few binding sites/fields further like the knight. And then it should be bounded by how far the knight can jump/ how many surface-bound adapter-oligos are in range of the first sequence. If you assume there are 2^35 in range there should not be a quadratic limit.
I guess the generated number depends on this "binding sites in range" for each sequence which is made up of "sequence/insert size" and "binding site density".
If this "binding sites in range" is like 4 you are right and it should be some kind of quadratical limit.
I don't think all the binding sites can be filled, no matter how many there are, because the existing bound "bridges" block each other. Even if they can criss-cross to some extent (and I'm not sure that they can), at some point they will physically impede each other from reaching free binding sites.