Troubleshooting Low Quality

DNA Sequencing Results

General Guidelines

* Avoid TE buffer and resuspend primers and templates in purified water

(10mM Tris-Cl, pH8.5 is ok)

* Avoid heat when drying samples - DNA can actually adhere to the plastic

tubes and hence be difficult to go back to solution

* Avoid freeze thaw cycles in templates over 9kb and avoid

drying these samples

* Make sure all phenol or chloroform, if used, is removed from samples.

* Primer design for sequencing:

Primers should be 17-24 bases in length

If possible, end with a 3' G or C, but no more than 2

GC content of 45-60%

Tm = 55 – 60.25 C calculated by (4(G+C)+2(A+T))

If possible, avoid homopolymer regions greater than 3 bases

Check primer dimmer and hairpin formation

------------------------------------------------------------------------

PCR Templates

* For the best results, make sure the templates run as one band and are

free of the PCR primers

* For complex PCR products, using sequencing primers that

are internal to the target template is recommended.

------------------------------------------------------------------------

Conversion Factors:

1 OD₂₆₀ unit = 33µg of DNA (Single Stranded DNA)

1 OD₂₆₀ unit = 50µg of DNA (Double Stranded DNA)

nmole DNA = (OD₂₆₀units x 90) / (length of DNA)

Important Note: Any OD reading outside the range of 0.1 – 1.0 is not reliable!

------------------------------------------------------------------------

Calculating the Molecular Weight of DNA:

MW = 312.2*#A + 328.2*#G + 288.2*#C + 303.2*#T - 61.1

------------------------------------------------------------------------

Calculating the Tm:

T_m (^oC) = 4.25 x (G+C) + 2.25 x (A+T)

T _anneal, (^oC) = Tm – 4.5

------------------------------------------------------------------------

Troubleshooting

A. If there is no reliable sequence due to failed reaction

1. Did you use a commercial kit or a home recipe to prepare your template? We strongly recommend Qiagen mini kits (Our experiences indicate that the maxi kits are not as reliable for DNA sequencing). If you used phenol-chloroform, did you use PEG for co-precipitation? Is there a contaminant? TE buffer? Phenol? Cesium Chloride? Chloroform? An ethanol precipitation can often eliminate these problems.

2. Check template and primer concentrations. What method was used to quantitate: by gel estimation, or by OD₂₆₀? Did you use reliable control for quantitation? If there is enough sample left, do the spec again. Try to provide the template and primer at the required amount.

3. Universal primers - Check vector map to make sure that the primer matches the template. A list of the bank primers and their sequences can be found in our web site:

http://www.people.virginia.edu/~dnaseq/primers.html

If it does match, is there another universal primer you can use? Often T7 and M13 are on the same side and T3 and M13 reverse. Sometimes the priming sequence on the template mutates and the primer won't work well. Furthermore, the template super coil can get nicked on the priming site thus rendering the primer unhybridizable.

4. User supplied primers - check Tm and %GC to comply with the good practice of primer design as mentioned in the general guidelines.

B. If reactions are working, but not well

1. Noisy baseline and low signals combined with Ns appearing in large numbers and starting early into the run: Are the signal strength numbers for the four bases under 50 (It is advisable to make it a habit to check these numbers, they are on the top right corner of the electropherogram. This can be caused by low amount of DNA template (check OD), priming problem (check primer design and Tm), or template purity (check for contamination by salts). Note that some templates are difficult to sequence because of their base composition, despite all the effort to perfect the run.

2. High signals- will often have a plateau at start but will drop off rapidly by 100 or 200 bases: Too much DNA - check OD and adjust OR dilute sample and run again. For short PCR products, changing the stop point can improve the balance.

3. "Hole" in a sequence - data drops off but then resumes: Check tracking. Make sure it is not off. If everything looks OK, then it is a gel compression, hence it is a bad gel. Run again.

4. Blip on gel, but no data: This is either a priming problem or due to low DNA concentration or contaminants. Sometimes 25pmol primer can solve (8x the normal amount). Try another primer if possible.

5. Stegosaurus data - 5-6 bases in a row alike then another base repeats: Looks like someone had a blast with the zig-zag feature of a sewing machine with different threads. The Taq has lost its footholds. This typically happens after a secondary structure that is particularly difficult or after a long poly nucleotide region. Often the only way to get data is to go from the opposite side and try to read through it. If it is a PCR product, be aware that the problem could have occurred during amplification and by putting it into a vector, the problem might be solved.

6. Signal drops after a poly nucleotide region (especially G's or T's) combined with multiple bases in each position: This is often the manifestation of the structural problems on the template. There are two options. Use a primer that sets on the poly N region. If you get through and know the final base that is great. Make the primer 17 like bases with the known base or make it a mixed base site if it is unknown, such as p(dT)₂₀N. The second option is to try from the reverse direction.

7. Noisy data, but signal is OK: If it starts after the first 60-100 bases, there is more than one product being sequenced, usually seen on plasmid, but, if more than one product was amplified by PCR you will see the same thing. If it starts immediately, the primer is priming to more than one spot. Often happens if primer forms a hairpin so only the first few bases bind to the template in numerous places. This requires a new primer to be made. If it is only in a few spots, you may have treaded upon some heterozygotes. This can happen if the DNA being sequenced is lethal to the host.

8. Good data in one strand but bad in the other, or, good data within vector sequence but run is bad when insert sequence starts: This is usually caused by double or multiple inserts, with a common 3’ or 5’ end, cloned into a vector. The remedy will be to re-pick the colony or even re-ligate the insert.

9. Looks like good data, albeit kind of weak, but unlikely to be mine: When your sample did not generate efficient amount of signals due to any of the aforementioned reasons, it is prone to be preached by the signals in its neighboring lanes, especially when the neighboring lanes have extraordinary signals. This has been a long existing problem yet to be overcome by the slab gel electrophoresis. By virtue, however, capillary gel electrophoresis is free of this trouble. To complicate the issue, some useful data appear weak and the investigators still value them. Therefore it would be too arbitrary for us to throw them away because of the weak signals. It has been our practice, therefore, to leave it to the investigator’s discretion.

------------------------------------------------------------------------

Final Notes:

We want to work with our users to provide them with the best sequencing service that we can. We will try to be here to help when there is a need. When a problem occurs, we would like to be a part of the solution instead of a party of the blaming game. So we encourage dialogs and discussions in which we are informed of the problem and offer our suggestions. Remember, the more information we have about your samples, the better we can modify reactions and troubleshoot any problems.

It is our general policy that there will be a charge for all reactions including those that do not yield useful data. When we detect any problems in our operation, we generally will rerun the samples involved and not deliver the questionable data. When a problem is brought forth, we will try to examine the issue in user’s perspective and often let the user have the benefit of doubt.

When a rerun is requested, the charge of the rerun is pending on the results. In case that there is significant improvement in the results, the rerun will be free of charge. On the other hand, if the rerun yields similar results, the cost of the rerunning will incur to the requestor.

We offer a trial period for new users so please inquire.