Skip to main content
Premium Trial:

Request an Annual Quote

Kennedy Krieger, U of Tuebingen Teams Develop New Tools for Chromosomal Microarray Analysis


Geneticists who use microarrays to detect chromosomal abnormalities have two new tools at their disposal.

One, called triPOD, was developed at Kennedy Krieger Institute in Baltimore, and allows users to determine the parental origin of an abnormality by comparing the SNP array data from a patient against data collected from both parents. The developers of triPOD recently discussed the tool in a BMC Genomics paper.

The other, called UPDtool, was developed by bioinformaticians at the University of Tübingen in Germany, and enables the detection and classification of uniparental disomy in SNP array experiments. The U of Tübingen team described the tool in the journal Bioinformatics.

BioArray News spoke this week with both teams of bioinformaticians about the new tools, and how laboratories conducting chromsomal microarray analysis could benefit from them.


The analysis tool triPOD was developed in the lab of Jonathan Pevsner at the Kennedy Krieger Institute. According to Pevsner, Joseph Baugher, now a postdoc at Johns Hopkins University and lead author on the BMC Genomics paper, drove the creation of triPOD while he was a graduate student in Pevsner's lab.

Baugher was studying SNP data from clinical cases and trying to find chromosomal abnormalities, said Pevsner, who is the correponding author on the new paper. "He wanted to detect deletions and duplications as well as copy number-neutral changes such as uniparental disomy," said Pevsner.

Additionally, Baugher wanted to determine the parent of origin for detected chromosomal abnormalities. "This is often of great interest in interpreting clinical cases," Pevsner said.

To address those needs, Pevsner's lab developed a tool that could compare SNP array data from an affected child against SNPs from both parents to report from which parent a particular aberration was inherited.

According to Pevsner, the tool identifies SNPs that are informative for abnormal parental contribution. As part of its analysis, it takes into account the genotype calls, the B allele frequency, and the log R ratio, related to copy number.

He said that triPOD includes modules for finding and defining so-called "streaks" of adjacent informative SNPs and defining their boundaries, as well as for detecting homozygous deletions, regions of Mendelian error, and cryptic regions having low information content.

He noted that its "sole limitation" remains its reliance on SNP array data from parent-child trios.

As part of its development, Pevsner's lab validated triPOD using a simulated dataset, and found it outperformed comparable software, including PennCNV, genoCNA, and bafSEG, for sensitivity of abnormality detection, and "displayed substantial improvement in the detection of low-level mosaicism while maintaining comparable specificity," according to the paper.

"The main advantage of triPOD is its excellent sensitivity and specificity," said Pevsner, claiming that "its sensitivity was dramatically better" that the other tools to which it was compared in the study. A figure comparing triPOD against other tools is available here.

The paper also showcases the tool's ability to detect low-level mosaic abnormalities from a large autism dataset across different Illumina microarrays. Because Pevsner's lab was able to detect low-level mosaicism using the trio data parsed through the new tool, Pevsner said that other researchers could benefit from using triPOD.

"So many biologically important chromosomal abnormalities occur as low-level mosaic changes," said Pevsner. "We found vast numbers of them in HapMap samples, in autism and other GWAS datasets, and in individual pedigrees," he said.


Like triPOD, the University of Tübingen's UPDtool relies on SNP array data collected from parent-child trios. But rather than focusing on parental origin of abnormalities, UPDtool detects both heterodisomic and isodisomic uniparental disomy.

UPD occurs when an individual receives two copies or a part of a chromsome from one parent and none from the other. It is said to be isodisomic when a single chromosome is duplicated, and heterodisomic when a pair of non-identical chromosomes are inherited from one parent. While isodisomic UPD is well characterized and considered to lead to the duplication of recessive genes containing potentially pathogenic variants, heterodisomic UPD is rarer and its implications are not as well understood, according to Christopher Schroeder, lead author on the new paper.

"Existing software packages can detect isodisomy stretches, but heterodisomy is more complex," Schroeder said. "You need trio information, and most software suites available are developed for single samples," he said. "As no one does this on a regular basis, so no one knows how frequently [heterodisomic UPD] occurs."

Schroeder and colleagues at U of Tübingen's department of medical genetics decided to create UPDtool because their department has a "large set of trios from patients with intellectual disability." He said that the lab was in need of a tool that was "easy to use and portable" and that could detect both kinds of UPD.

"The previously existing tools did not meet our requirements, especially towards usability and ease of installation on all major operating systems," Schroeder said.

As detailed in the paper, the resulting algorithm was tested using five positive controls including both iso- and heterodisomic segmental UPDs and 30 trios from the HapMap project as negative controls. With UPDtool, the authors were able to correctly identify all occurrences of non-mosaic UPD within their positive controls, while no occurrence of UPD was found within their negative controls.

In addition, they reported that chromosomal breakage points could be determined more precisely using UPDtool than by microsatellite analysis, which is currently the standard for detecting UPD. Schroeder and colleagues also benchmarked UPDtool against another software tool for UPD detection called SNPtrio, which was developed by and hosted at Pevsner's lab at Kennedy Krieger.

Schroeder said that his department intends to use UPDtool in regular diagnostic screening, in order to develop a better understanding of how heterodisomic UPD could contribute to the transmission of pathogenic chromsomal abnormalities.

"UPD, especially heterodisomy, is difficult to detect, and screening for UPDs is rarely done in routine diagnostics," Schroeder said. "We are planning to implement it to see how many cases of UPD we can find," he said. Schroeder added that they will focus on hereditary diseases where the causes have not yet been identified.

Meantime, the U of Tübingen team is also making enhancements to UPDtool. Bioinformatician Marc Sturm told BioArray News that they are making small changes to the algorithm to detect smaller stretches of UPD as well as making it easier to install and developing its graphical interface.

He also said that array vendors could consider adding such features to their existing software analysis tools.

"If you have the array information, it's not too complex to detect UPD," said Sturm. "Vendors could add that capability."