GM_ICW Table

The GM_ICW table records GEDmatch in-common-with (ICW) relationships — pairs of kits that both match each other. Each row links two kits and records how much DNA they share. This is the data behind clustering: knowing which of your matches also match one another.

This page documents the schema and the behavior of the existing gather code as written.

kita and kitb

kita and kitb are the two kits in the in-common-with relationship. As elsewhere in the GEDmatch tables, kita is treated as the primary side and kitb as the other kit. ICW rows are written bidirectionally — each relationship is inserted twice, once as (kita, kitb) and once with the two swapped — so the relationship can be queried from either kit’s perspective.

No explicit primary key. Unlike the other GEDmatch tables, GM_ICW has no Id / primary-key column. Rows are de-duplicated at insert time (insert-or-ignore on the kit pair) rather than keyed by a surrogate id.

How It Is Populated

ICW rows come from the ICW / triangulation stage of a GEDmatch gather and from AutoSegment link data (the “links” between nodes in a cluster graph). The cM value is read from whichever shared-cM field the GEDmatch response provides (total_cm, totalCM, shared_cm, or cM); AutoSegment links are stored without a cM value.

Column Reference

Column	Type	Description
kita	string	The primary kit in the in-common-with pair. References `GM_Kit.kit_num`.
kitb	string	The other kit in the pair. References `GM_Kit.kit_num`.
totalCM	float	Total shared DNA between the two kits, in centimorgans. 0 when the source (e.g., AutoSegment links) does not provide a value.

Related Tables

GM_Match — the match summary; its icwdate is set when ICW data is gathered for a pair.
GM_Kit — resolves kita / kitb to people.
GM_TriangulatedSegment — triangulated clusters built from overlapping shared segments.