Tree illistration,
April 23, 2024 | Human ID and Forensics

Searching for Relatives in GEDmatch

A comparison of ForenSeq Kintelligence and whole genome sequencing
GEDmatch is an important investigative tool for law enforcement, particularly for cold cases and unidentified human remains. In order for law enforcement to compare samples to the GEDmatch database, several types of sequencing data can be uploaded to GEDmatch PRO, including microarray, whole genome sequencing (WGS) and data generated by the ForenSeq Kintelligence Kit.

My colleagues and I wanted to provide a comprehensive overview of how WGS and Kintelligence compared using a known pedigree with distant relationships. We compared the performance of WGS and Kintelligence in identifying true relationships in GEDmatch PRO. Our results allow investigators to make an informed decision about a sample based on the availability of funding and access to expertise.

1. What were the comparison criteria?

Using WGS and Kintelligence, we sequenced the DNA from a member of a large family that contains up to 6th degree relationships. Both degraded and intact DNA samples were included with inputs of 1 ng and 100 pg. For WGS data, we utilized the One-To-Many Segment Matching tool in GEDmatch PRO to search for relatives in the pedigree. For Kintelligence, we utilized the One-To-Many Kinship tool.

2. What did the comparison reveal about WGS and GEDmatch PRO?

We demonstrated that WGS processing requires a large amount of bioinformatics expertise. Each sample required a different bioinformatics workflow to maximize true positives and minimize false positives. This approach is not reproducible and not ideal if standard protocols must be in place for sample processing. Furthermore, although WGS returned a high number of SNPs typed, only a few of the known relatives were matched.

3. What did the comparison reveal about Kintelligence and GEDmatch PRO?

The Kintelligence workflow required no bioinformatics expertise. The same workflow was applied to every sample because all analysis is contained within the Universal Analysis Software (UAS). All samples returned >9000 SNPs typed (of 10230 total SNPs targeted in the assay) and matched all known relationships up to 4th degree. In addition, the intact 1ng and 100pg samples matched one of the two 6th degree relationships.

Overall, the Kintelligence workflow is reproducible, was easier to use and returned the highest number of true positives and very low false positives. Disease and/or medically informed SNPs are excluded from the Kintelligence assay whereas these markers must be removed bioinformatically from WGS data.

4. Is WGS the best option for samples with low or degraded DNA?

Some investigators have heard that WGS is the best option for working on cases with samples that have very low amounts of DNA and/or the DNA is degraded. The assumption is that WGS will get the most out of this type of sample, that with more data, more true matches will be returned from querying GEDmatch.

5. Is this a false assumption?

The amount of bioinformatics required to process a WGS sample is often overlooked. Performance of WGS data in identifying true relationships in GEDmatch PRO is not well-established and often comes from word of mouth.

6. Overall, does Kintelligence outperform WGS in the search for relatives in GEDmatch PRO?

Yes, Kintelligence was designed specifically for forensic samples, has a streamlined workflow that requires no bioinformatics knowledge. Given high quality data, Kintelligence can match up to 5th degree relationships.

7. What do you see in your scientific future?

As an employee in the R&D department of QIAGEN, I will be contributing product development of new, cutting-edge products and continuing to improve our existing products, including GEDmatch PRO.


Comparison of ForenSeq® Kintelligence and Whole Genome Sequencing in Searching for Relatives in GEDmatch PRO™


S.M. Radecke1, J. Antunes1, G. Padmabandu1, K.M. Stephens1  Affiliation: QIAGEN, San Diego, CA USA