PVS1
This deep intronic duplication does not fall within the PMS2 VCEP default PVS1 categories for nonsense, frameshift, canonical splice, or qualifying large rearrangement variants, and no patient RNA evidence was identified to show that the variant causes a loss-of-function splicing abnormality.
PS1
No evidence was identified that this variant affects the same non-canonical splice nucleotide as a previously established pathogenic or likely pathogenic PMS2 splice variant with similar or worse predicted splicing impact.
PS2
No de novo data with the required tumor context or parental confirmation were identified, so this criterion cannot be assessed from the available evidence.
PS3
No calibrated functional assay or constitutional RNA study demonstrating a damaging effect was identified for this variant.
PM2
The gnomAD v4.1 total allele frequency is 5.0248e-05 (81/1612004 alleles; 0.00502%), which is above the PMS2 VCEP PM2 threshold of 0.00002 (1 in 50,000 alleles), so PM2 is not met.
PM3
No observations were identified to score this variant in trans with another pathogenic PMS2 variant for the PMS2 VCEP PM3 point system.
PM5
This is not a missense variant, and no evidence was identified to support applying a same-residue missense rule.
PP1
No segregation data were identified, so co-segregation support cannot be assessed.
PP3
SpliceAI predicts a maximum delta score of 0.02, which is below the PMS2 VCEP PP3 threshold of 0.2 for a predicted splice defect, so PP3 is not met.
PP4
No MSI-high tumor findings or mismatch repair protein loss data consistent with PMS2 were identified, so phenotype-based support is not established.