Home
Services
Pricing
FAQ
Contact
Philosophy
About

Researcher Login Enter username and password to access your results pages:

Username:

Password:

---

Frequently Asked Questions

1. What is DNA pooling?
2. Why DNA pooling?
3. Why not DNA pooling?
4. How big can my pools be?
5. What about confidentiality?
6. What about intellectual property?

---

What is DNA pooling?

DNA pooling is a technique, in which DNA samples are combined together and assayed as a single sample. For quantitative genotyping assays, the resulting genotype information can be used to estimate the allele frequency of the sample. Using this estimate, allelic association can be tested using a modified Chi-squared statistic.

Why DNA pooling?

DNA pooling has two main benefits - cost and data efficiency. By combining samples into pools, the cost is lowered in proportion to the size of the pools. That is to say that is you have 100 cases and 100 controls and group them in pools of 10 samples in each pool, then the cost will be roughly 10x less expensive than if you were to run individual samples. There is, however, some overhead for each sample and, for example, combinations of 10 individual samples evaluated as a single pool would be 5x less expensive.
Reducing the data load by an order of magnitude makes some much more time intensive algorithms practicable. Since the size of each input is already on the order of 10⁴ or 10⁵ genotypes, any reduction in data load is helpful to the investigator. Sorting through a list of those differences 3 standard deviations higher than the mean becomes a daunting task when there are over 1000 SNPs on the list.

Why not DNA pooling?

There are three reasons not to use DNA pooling: error rate, information loss, and cost.

1. Error rate
   The pooling step introduces another source of error into the genotyping experiment. Some error in pooling is unavoidable and running individual samples will, in general, produce more robust results. We have reduced the standard deviation to an acceptable level ~5% and are working every day to further reduce the error. By dividing the samples into multiple subsets and comprising the pools from each subset the error is reduced from the ~5% by (n-1)^1/2 where n is the number of subsets or pools. So using 10 subsets would provide a standard error of the mean of 1.7%. An error this low means that results showing a 10% change between cases and controls are usually highly significant.
2. Loss of information
   Technically speaking, DNA pooling experiments produce allelotypes. That is, they generate only allele frequencies and certain information about genotypes and haplotypes cannot be obtained using our current pooling methods. This is the flip-side of the data reduction coin.
3. Cost
   If money is not an object (and for some extremely well-funded investigators, it is not), then DNA pooling is contraindicated. If you can afford to run you thousands of samples on microarrays that cost thousands of dollars a piece, then by all means do so! For most of us, however, multi-million dollar budgets are not the order of the day.

How big can my pools be?

In the past we have had success using pools as small as 6 individuals and as large 50. With more experience this range may expand or contract. One bad DNA sample can lead to poor data recovery from the array.

What about confidentiality?

Your results stay within the company until you are ready to let us talk about them. They are placed into the JKA database where only JKA employees and JKA consultants have access to them.
We plan to use your data for the improvement and development of JKA products and services and after dissociation from phenotype to aid in the analysis of data from other customers and collaborating investigators.

What about Intellectual Property?

The customer owns the intellectual property developed from their samples that is relevant to their phenotype of interest.