Table 1.

Summary of CDS orthology between HAL2 and FIL2

Category (HAL2: FIL2)	n. genes	% Low support transcripts	% Expressed F2 leaf
1:1 Orthologsa	34,396	7.8	74.6
1: 2+ Orthogroupb	548	36.7	42.3
2+: 1 Orthogroup	484	39.0	38.6
2+: 2+ Orthogroup	5589	23.2	51.8
Privatec to HAL2	4455	77.7	20.6
Private to FIL2	4971	73.0	19.7
Presentd: un-annotatede	160	84.3	7.3
Un-annotated: present	136	88.6	4.6
Present: pseudogenef	209	38.0	36.5
Pseudogene: present	212	38.4	43.7
Present: deletedg	19	52.6	35.7
Deleted: present	14	57.1	38.9

Orthology, un-annotated, and presence/absence are inferred via the GENESPACE pipeline (Supplementary Note 3). The gene-annotation category (where HAL2 category precedes ‘:’ and FIL2 follows) and number of genes found therein are presented in the first two columns. Annotation confidence score (Supplementary Note 2) is calculated via the degree of homology, gene expression, and PFAM support. Gene models that did not satisfy these criteria were considered low support. To test for expression in leaf tissue in our field experiment, we counted transcript abundance in an HAL2-FIL2 F₂ population. Expressed genes had at least one count in ≥10% of the F₂ population and mean counts >5, after excluding libraries with 0 counts.

^aOrthologs are pairs of HAL2 and FIL2 gene models, where a single gene model from each genome is represented in an orthofinder orthogroup.

^bOther orthogroups contain two or more gene models from one or both P. hallii genomes.

^cPrivate genes are found in single-gene orthogroups without representation of any outgroup (S. viridis and S. bicolor) sequences.

^dPresent genes are found in orthogroups that contain one or more outgroup genes, while ^eun-annotated genes have sequence with >90% coverage in the alternative genome assembly but no gene annotation.

^fPseudogenes have >10% and ≤90%, and ^gdeleted sequences have ≤10% similar sequence coverage in the alternate genome assembly.