waigitdas
commented
3 years ago Currently TGs are manually curated and delivered to java code via a csv file, using the template provided. A UDF loads this, creating tg nodes and edges to kits, persons and segments. Another UDF sets a specified ancestor_rn into kit, person and DNA_match nodes, then finds all descendants who have DNA test results and then their mapped chromosome segments and finally creates an edge linking them in their location on the chromosome (seg_seq edge).
A few UDFs create reports on TGs which include a list of known matches, common ancestors, ahnentafel paths and relationships. The documentation for these needs to be written.
Triangulation groups are fundamental in genetic genealogy analytics. Ideally, they would be automatically identified. That is a challenging task in a multiple kit environment because of the nearly continuous overlap of shared segments which makes boundary detection difficult. Thus, an output which allows the end user to see matches and segments will allow them to curate the boundaries. These curations need to be organized in a text document which can be consumed in analytics.