Protocol Online logo
Top : Forum Archives: : DNA Methylation, Histone and Chromatin Study

How to interprete direct sequencing results? - (May/19/2007 )

Pages: 1 2 3 Next

Hi,

after some work, now I have some data to evaluate, but.. what is what?
From DNA extraction to gel-extraction after BSP i worked totally with Qiagen-products and i sequenced (direct sequencing) with the reverse Primer. Now I observe that in the majority of the samples the G -basic-signal is little unstable so its difficult to differentiate between a possible methylated C and/or an artifact (see attachment). Furthermore I observed that some Probes shows a relatively high number of GC´s (with a very strong G-signal) wich means a methylated GpC`s on the sense strand (???) ..that gives no sense..
Is it maybe a question about DNA -purity before BS-treatment, or can I interprate the G-signal as a in that cell-group partially methylated CpG?

Thanks for the reply,
Tharom

[attachment=3011:Sequence.gif]

-Tharom-

Hi Tharom,

My impression is that there is likely partial methylation at those "N" positions. Incomplete conversion is apparently a problem because the G noise seems to only appear under "A" peaks. If a sample doesn't have any methylation, "G" signal may appear as random background. Yours may be a different case. Overall, your sample has little methylation. If I were you, I wound stop here and designate this sample as having no methylation.

-pcrman-

I second PCR man on that issue. Still, I think you have a problem with incomplete conversion. This could interfere with your results, even in more methylated samples. I suggest that you try to perform some mixing experiments with completely methylated and unmethylted DNA to check your bisulfite and measuremnts efficiency.

-krümelmonster-

hey Tharom,

can you check what the signal intensities are for each base signal? This can be found be opening your ab1 file and having a look at the additional information and under SIGN you have the numerical values for each base, see in my situation (it's a clone) the G signal is much lower than the others (see attachment), if you have a weak overall signal from your direct sequencing what you will find is that the bascaller will try and overcompensate for this and hence the noise.

I would be interested to hear what your values were and if you were using the ABI to sequence.

cheers

Nick

-methylnick-

Hi Nick,

is SIGN the average raw signal intensity? In ABI Sequence Scanner this can be found on the annotation sheet.
The attached is from direct sequencing of a unmethylated control. The traces were more or less free of G backrgound, so I think raw intesities above 100 should work.

Krümel

-krümelmonster-

Hi, thanks for the help!

Methylnick: I looked for the basic signals, but the G-basic signal is very low?? [attachment=3028:Seq.gif]. Yes, the probes were sequenced with an ABI (=some (dis)-advantages?)

Probably it is a DNA-purity problem: Today I got the results form other Probes who were stored before Bisulfite in a RNA/DNA stabilization reagent (Roche) who lyses the cells. This probes shows a much better sequence and only a G-ground-signal.. rolleyes.gif

Cheers Tharom

-Tharom-

@krumel, yes that is the number i am talking about. and intensities above 100 are okay.

@tharom, those signals are very very low!!!

Nick

-methylnick-

Hi Methylnick,

I´m a newcomer in the research and I have not really experience with sequencing, but those values doesen´t make any sense for me.. I´mean, the electropherogramm shows a clear, irregular and sometimes very strong G-signal and in the additional information i get a value around 10 (???).
I´m without ideas...

Tharom

-Tharom-

Hi Tharom,

what you see in the electropherogram is what the software makes of your raw values. The raw signal intensities of your reading are very low, so the software uses it's routines to make educated guesses. This leads to a lot of irregular G-spots, as it tries to average the signals. If you use Sequence Scanner by ABI you can overlay raw signals and analyzed signals - there you get an idea, why G's may appear.
This software adjustment is a big problem when you have very low intensities for all signals. It is much better, if you have better signals in all four channels. You can also trick out the software by using C or G enriched primers, which partially correct the readings. The ESME software by epigenomics uses the raw values from ABI sequencers and after a lot of transformation, methylation information is presented.
But for you the essential should be to improve the general signal intensities. How much PCR amplicon do you use for sequencing?

Cheers,
Krümel

-krümelmonster-

Hi Krümel,

interestingly in some samples with the same signal-values I get acceptable electropherogramms, [attachment=3029:Dokument1.pdf]. (?)
A company sequence our probes and they sequence with 7-8 ul amplicon and they need 10 ul primer at a concentration of 10pmol/ul. The run in the attachment shows 3285 scans = amplicons (?) Is it possible to get better results when I increase the primer concentration for sequencing?

..so if I understand it right, in my case, the irregular G-signal is more likely to explain with a sequencing problem as with a incomplete DNA-conversion trough bisulfite ?

Thanks
Tharom

-Tharom-

Pages: 1 2 3 Next