I have a primer than can prime several genomic areas at once (that's the intention). The number of sequences I get back for an individual locus which is uniquely aligned varies however. I'm interested in seeing if there are any sequence specific factors that might be responsible other than the primer matching imperfectly or GC content. Does anyone know what DNA specific factors can affect PCR efficiency?
Just so I'm following you here Sebastian, your primer primes several locations in the genome - resulting in fragments within the range sequencable by the sequencer?
Because if the answer is no, then you have problems. For example, if your primers bind but the resultant fragment is too big/small, you will have this huge population of fragments in your library of which the sequenced DNA which ends up smaller (random) % of. In other words, the counts you see for your reads at one locus will be effected by the known counts of another locus (sequenced, deliberately) BUT ALSO THE UNKNOWN COUNTS of the fragments that were never sequenced because they were too long/short.
If you don't want to be quantitative, that's fine - you can get a true/false value for regions primed.
Other factors which effect PCR efficiency include:
3-dimensional folding of the AMPLICONS (check in a ssDNA folding program like mFold) - this is only a factor if although your primers prime regions of a similar size/GC, the sequence of the amplicon is different, resulting in a different 3D structure during cooling, and if there are hairpins in the primer binding sites in some of your amplicons their PCR efficiency can be significantly lower.
GC content of the DNA around the target site, even if the target site has a similar GC%
The fragments are size selected prior to being placed on the sequencer (gel extracted) so I don't get problems such as the one you mentioned. It is a very narrow fragment size for all loci.
I have a number of amplicons that vary slightly- by about 11% in a 150bp region. I'll try to run mFold and see if this variation has an effect.
Regarding the primer binding site this shouldn't be an issue because in the analysis I only select sequences that match the primer exactly.
So I guess the two factors are GC around the target site and amplicon folding? Do you have an idea what distance around the target site may mean? Is there any consensus on this from previous studies?.
Sorry my first message wasn't very clear. Reading it back myself I don't think I explained it very well...
Even if your primers bind perfectly/identically, the 3D structure of your amplicons depend on their sequence. As an example, one amplicon might have the reverse-compliment sequence of your primer in the middle of it, while another doesn't. Thus, during PCR, the first amplicon will amplify very poorly, while the second amplicon will amplify very well - even though the priming sites are identical. In the first amplicon, your primer cannot bind because the site is busy interacting with the amplicon itself.
The only way to check is to use mFold (or something similar - I haven't used mFold in like 3 years...) to see if you have hairpins in the primer binding regions :)
You... probably will... at different free energy values you'll have aaaall kinds of structures. Whether those isoforms exist in real life is another question. But it is very likely to be the reason you see different amplification rates of your amplicons.
ADD REPLY
• link
updated 24 months ago by
Ram
43k
•
written 9.2 years ago by
John
13k
OK. Thanks John
The fragments are size selected prior to being placed on the sequencer (gel extracted) so I don't get problems such as the one you mentioned. It is a very narrow fragment size for all loci.
I have a number of amplicons that vary slightly- by about 11% in a 150bp region. I'll try to run mFold and see if this variation has an effect.
Regarding the primer binding site this shouldn't be an issue because in the analysis I only select sequences that match the primer exactly.
So I guess the two factors are GC around the target site and amplicon folding? Do you have an idea what distance around the target site may mean? Is there any consensus on this from previous studies?.
Cheers
Sorry my first message wasn't very clear. Reading it back myself I don't think I explained it very well...
Even if your primers bind perfectly/identically, the 3D structure of your amplicons depend on their sequence. As an example, one amplicon might have the reverse-compliment sequence of your primer in the middle of it, while another doesn't. Thus, during PCR, the first amplicon will amplify very poorly, while the second amplicon will amplify very well - even though the priming sites are identical. In the first amplicon, your primer cannot bind because the site is busy interacting with the amplicon itself.
The only way to check is to use mFold (or something similar - I haven't used mFold in like 3 years...) to see if you have hairpins in the primer binding regions :)
You... probably will... at different free energy values you'll have aaaall kinds of structures. Whether those isoforms exist in real life is another question. But it is very likely to be the reason you see different amplification rates of your amplicons.