Protocol Online logo
Top : Forum Archives: : Molecular Biology

Sequencing non-existent DNA? - (Nov/01/2008 )

I am trying to sequence a gene in a human, but I am getting very strange results... on 3 of the 4 exons that have come off the machine (ABI310) I have managed to sequence DNA that does not exist in any animal (according to NCBI BLAST). The sequencing trace is very clean and the one exon that didn't have this problem worked well. The DNA that I did manage to sequence hits to some part of the human genome (albeit different chromosomes for every odd exon) however, the sequence right before the end of the sequence only matches the primer for 5-10 of the 3' most bases. For example, I might have some sequence from a different chromosome, but the last 20 bases are of the primer, but only the 10 3' most bases match the rest of the sequenced DNA (the other 10 do NOT match the unexpected product whereas they DO match my reverse PCR primer which matches the gene I want to sequence). It's almost like my primers are only annealing only on the first 5-10 bp... how/why is this happening and I still have good sequence?

-eascsc-

This happend to me once, completely non-existent sequence. I found I used the wrong basecall module in plate setup, so all Cs were Gs and so on. But this isn't the case if you say you have the right sequences of the primers, but only on small portion of it.

So..
are you sure your primers don't match repeat or low complexity sequences?
isn't the primers sequence somehow screwed?

If you say "but only the 10 3' most bases match the rest of the sequenced DNA" you mean 3' of the sequence or 3' end of the reverse primer?
Like this:

CODE
sequence   5' ATGTAGATAGATGATAGATA 3'
                  |||||||||
primer         3' TCTATCTACGGCACCC 5'

or like this:
CODE
sequence    5' ATGTAGATAGATGATAGATA 3'
                         ||||||||||
primer           3'GCCTATTACTATCTAT 5'

-Trof-

The problem appears to be the first of your two examples

CODE
sequence   5' ATGTAGATAGATGATAGATA 3'
                  |||||||||
primer         3' TCTATCTACGGCACCC 5'


But if this is the case, the first 10 bases would have to occur at many of spots around the genome, no? If so, then how can only one of those many products be sequenced?

-eascsc-

QUOTE (eascsc @ Nov 2 2008, 12:54 AM)
But if this is the case, the first 10 bases would have to occur at many of spots around the genome, no? If so, then how can only one of those many products be sequenced?

They may, or there may be a pseudogene of your gene of interest. Or there may be one repeat so abundant, that would amplify preferentially.
There are many options what could go wrong, but none would lead to a nice sequence of something completely different. Did you check your machine and settings?
Anyway, do you wish to post your primers sequences (one problematic pair for example)? Maybe that could tell more.

-Trof-

Here are the primers (designed for a Tm of 60*C by ExonPrimer, PCR annealing step performed at 59*C).

F-ctcagggtctgcatgtttcc
R-tcacatgcccagctactcag

The gene I am sequencing is EDEM2. Here is end of part of the sequence of what actually came off of the machine:
TGTGAGCTCAAGTGATTCTCCCACCCCAGCCTC[CTGAGTAGCTGGGCATGTGA]- my reverse primer (well, the reverse compliment of the reverse primer because I sequenced with the forward primer) is in brackets. If that whole sequence is BLATed, there is no such sequence that actually has the full reverse primer in it. I checked the machine with a temperature probe and with primers that I know worked and I got perfect sequence of what the primers were designed to amplify and the product was of 1 clean band on a gel.

-eascsc-

Just from the first blasting of your forward primer it shows multiple high-probability hits on many chromosomes.

I quick test the part of you sequence in the Repeat masker and it showed an Alu repeats on 26% of your amplicon. You may check it yourself and redesign your primers. This may not really explain your weird results, but those primers are definitely bad.

Always blast your primers before ordering to see if they don't hit a repeat. Or use Primer3 with selected Repeat library to prevent designing a primer in repets.

-Trof-

QUOTE (Trof @ Nov 3 2008, 12:51 AM)
Just from the first blasting of your forward primer it shows multiple high-probability hits on many chromosomes.

I quick test the part of you sequence in the Repeat masker and it showed an Alu repeats on 26% of your amplicon. You may check it yourself and redesign your primers. This may not really explain your weird results, but those primers are definitely bad.

Always blast your primers before ordering to see if they don't hit a repeat. Or use Primer3 with selected Repeat library to prevent designing a primer in repets.



That's weird because I used ExonPrimer which uses the Primer3 algorithm (I always select the Human mispriming library with default settings for the other mispriming parameters) ... I used to make all my primers by hand until recently and I have had a couple primers that ExonPrimer made that just plain don't work (at that point I re-make them by hand). I guess I should give up on ExonPrimer and make all of my primers by hand.

Thanks!

-eascsc-

Maybe this helps you: Chimera check

A programm to identify chimeric sequences; maybe something goes wrong during your PCR?

-gebirgsziege-

Have you tried NCBI Primer BLAST?

http://www.ncbi.nlm.nih.gov/tools/primer-blast/index.cgi?

Using your primers and selecting the db as human genome reference sequence it appears your primer hits a single location on Chromosome 20 human. This si a good sign at least...


Primer pair 1
Sequence (5'->3') Length Tm GC%
Forward primer CTCAGGGTCTGCATGTTTCC 20 60.66 55.00%
Reverse primer TCACATGCCCAGCTACTCAG 20 60.01 55.00%
Products on intended target
Products on allowed transcript variants
Products on potentially unintended templates

>NT_028392.5 Homo sapiens chromosome 20 genomic contig, reference assembly

product length = 480
Forward primer 1 CTCAGGGTCTGCATGTTTCC 20
Template 3915815 .................... 3915796

Reverse primer 1 TCACATGCCCAGCTACTCAG 20
Template 3915336 .................... 3915355


-Watti-

QUOTE (Watti @ Nov 5 2008, 08:09 AM)
>NT_028392.5 Homo sapiens chromosome 20 genomic contig, reference assembly

product length = 480
Forward primer 1 CTCAGGGTCTGCATGTTTCC 20
Template 3915815 .................... 3915796

Reverse primer 1 TCACATGCCCAGCTACTCAG 20
Template 3915336 .................... 3915355

[/font]


Here is the actual region on the mapviewer...

http://www.ncbi.nlm.nih.gov/projects/mapvi...355&thmb=on

-Watti-