I have been recently asked to assess what is the better technology for screening for genetic diseases and why would someone want to use microarrays to do it. NGS and microarrays are very different technologies with their own pros and cons. So let’s look at what they are and which one is truly better.
Targeted re-sequencing applications analyze a subset of human genes and/or mutations. Selection of the genes varies from project to project. Recently the biggest interest is in using targeted sequencing in human genetics, in particular in carrier screening and diagnosis of genetic diseases. There are multiple companies trying to get into the screening market as prenatal screening and carrier screening are becoming more commonly used.
Samples are primarily selected by targeted PCR amplification. Amplicons can then be used in both microarrays or in NGS. However, the amplification can induce bias, errors and allele drop off due to misspriming or lack of primer binding due to SNP presence. Alternatively, NGS applications use probes for selection of genes of interest. Subsequent PCR then uses non-specific sequences for amplification. Is this a better approach? Not necessarily, there is possibility of missing the target if there are mutations in the sequences used for capture. So while the errors would be different then those introduced by PCR they can still be present. The real difference is the downstream technology and data processing.
Microarrays rely on the Affymetrix technology for sequencing by hybridization. This technology allows to read sequence by comparing the reference sequence on array to the sequence in a sample. After the sample is hybridized to the array, signal is converted to nucleotide sequence, and this sequence is then compared to the reference on the array. The result shows you which mutations are present in the sample. Up to last year there were multiple tests using this technology – Cardiomyopathy panel (Harvard Partners), Noonan Syndrome panel (GeneDx) and Pan-Ethnic carrier panel (TessArae). Currently, there is only one genetic re-sequencing assay on the market – Pan-Ethnic carrier panel (90 genes and over 400 mutations), the other two have been replaced with NGS panels.
NGS targeted re-sequencing has been developing for the past few years. There are many panels that were developed by individual testing labs, such as GeneDx, but there are also those that are being developed by companies such as Illumina. Collaboration between Harvard Partners and Illumina generated a cardiomyopathy panel (46 genes). In addition, Illumina has developed an inherited disease panel (552 genes), cancer predisposition panel (94 genes) and an exome subset containing clinically actionable genes (2761).
What are the pros and cons of the two technologies?
Microarray Affymetrix technology is well known, there are many users familiar with the workflow and nearly instant analysis of results means that technology is easy to implement. Sample to results time is rapid (24 to 48 hours), accuracy is comparable to sequencing. However, there are limitations, it has been known for some years now that any changes that are not base substitutions an be difficult to identify reliably using re-sequencing arrays. Yet, it is not a consistent problem, some of those changes are easily and reliably identified but it takes some validation work to prove it. There is also no way of identifying de novo mutations which are not base pair substitutions, which can be a problem if the genetic disease results from de novo mutations. The issue that can be a pro or con is a relatively low throughput. Small labs will like the ability to scale it as they need, large labs can find this to be a problem.
NGS targeted sequencing makes analysis relatively easy, the sample preparation is also much faster now, allowing the sequencing to be done in 24-48 hours. As it is sequencing, identifying any de novo changes is possible. The problem posed by NGS is still the cost of the test, which can be kept down only by running many concurrent samples, including multiplexing them. NGS is definitely a high throughput method. A significant issue with NGS can be pseudogenes and repeats, but then microarrays are not immune to those problems.
Right about now you want to hear which tool to use. The choice is going to based on your lab needs, experience and infrastructure. If you want high flexibility of the assay, then NGS is probably the most sensible route as it allows you to make changes at any stage as you need them. If the disease/s you are testing can have large number de novo mutations and/or insertions/deletions/indels then again NGS is probably going to be a method of choice. However, if you want a quick, very easy assay testing for primarily nucleotide substitutions, then microarray method would be a good choice. An existing Sanger sequencing assay can be converted into either of those two with some effort. All that will be needed is a custom analysis method for the results. One additional consideration is the cost of validation of the method before it becomes a clinical method and can be the decision maker for your lab.
Leave a comment