Table 1.

EPHB1 mutations identified in 9,898 cases of 33 different TCGA tumour types. TCGA pan-cancer pan-EPH somatic mutations were filtered and EPHB1 coding non-synonymous mutations with a high functional impact were selected if they either had (i) another cognate mutated amino acid in close proximity in 3D-space or (ii) by mutation prevalence. Primary in vitro phenotypical screening by IncuCyte identified the mutations with compartmentalisation phenotype for validation by confocal microscopy. (*) Mutants selected for confocal analyses. The position (a) V248M is located between the ligand-binding and the fibronectin type-III 1 domains in a compositional bias region enriched for cysteine; and (b) R883Q is located in the position C-terminally to the protein kinase domain. n/a, mutated positions were not associated with another amino acid in the 3D space

EPHB1 Amino acid Position	EPHB1 Mutations	Total EPH Mutations	3D-Partner Position (EPHB1)	Protein Domain
G685D*	1	4	I696	Protein Kinase
R743W*	5	13	F801	Protein Kinase
G821R*	2	7	R170	Protein Kinase
C61Y*	1	2	C183	Ligand Binding
R90C*	2	6	S188	Ligand Binding
R682C	3	10	E698	Protein Kinase
T117I*	2	3	G172	Ligand Binding
R170W*	7	9	E116	Ligand Binding
V248M	1	26	n/a	(a)
R883Q	1	18	n/a	(b)
R748S*	2	14	n/a	Protein Kinase
D762N*	6	14	n/a	Protein Kinase
R865W*	3	13	n/a	Protein Kinase
R351L*	2	11	n/a	Fibronectin type-III 1
R799H	1	10	n/a	Protein Kinase