CLUSTER 3, with its modal values shaded in blue, is too distant genetically from Clusters 1 and 2 to have shared a common ancestor with either since hereditary surnames came into use. Eleven of Cluster 3’s twelve samples have been tested at 37 or more markers, six of them at 67 markers. Five samples are from descendants of a family living at Kirkneedy, near Letterkenny, Conwal Parish, County Donegal, by 1800. Another is from a family living about the same time in Ballynakill Parish, County Galway, near Loughrea (like an unrelated sample in Cluster 2). The Conwal and Ballynakill Y-DNA samples match exactly at 25 markers, and differ by only two steps at 37 markers, showing that their common ancestor was quite recent. A match that close and sharing the same surname indicates that he probably lived within the last six generations, but no records go back far enough to confirm the relationship. Two of the 37-marker Conwal samples differ by only two steps from the Ballynakill sample, the same genetic distance that separates known first cousins in the Conwal line.

A  sample from a family first of record in Kilconduff Parish, County Mayo, in 1826, matches the Donegal and Galway samples at 33 of 37 markers. Two other samples, from families first found in the US by 1750, match each other at 35 of 37 markers, and are within three to five steps of the modal values, showing a relatively recent Devine ancestor in common.

Cluster 3 is so genetically distant from Clusters 1 and 2 that it has a quite separate ancestry and probably a different Irish locality of origin, perhaps representing the Ó Daimhín sept recorded in County Fermanagh from the 1100s to the 1400s but now seldom found there after being displaced by the Maquires.

CLUSTER  4 (yellow shading) is similarly too distant genetically from the other groups to share a common Devine ancestor. The four samples represent related lines, two rather close at 37 markers, but the other two, with a greater number of mutations from the cluster modal values, showing that their common ancestor lived in the more distant past.

UNCLUSTERED (gray shading), are five samples. Three of them, like all those in Clusters 1, 2 and 3, appear to be from the large Haplogroup R1b, to which most Irish males belong. The families are first of record at unspecific Irish locations and Tennessee, USA, before 1850. They are genetically distant enough from each other and from the three group clusters that their common male ancestor would have lived well before surnames came into use. The final two ungrouped samples belong to a completely unrelated population group that arrived later in Ireland, Haplogroup I. Found extensively in Scandinavia, in Ireland it suggests a Viking or Norman origin. One sample is from a family first of record in present West Virginia, USA, in 1806; origin of the other is not yet reported. Neither line is related to the other Devines because they are in different major haplogroups.