This variant is a missense substitution (c.2395C>T, p.Pro799Ser) in exon 20 of DNMT3A (NM_022552.4). It is absent from ClinVar and has an extremely low allele frequency in gnomAD v4.1 (4/1,613,986 alleles; AF=0.00025%), meeting PM2 (supporting). In silico prediction tools support a deleterious effect (REVEL=0.916; BayesDel=0.504), meeting PP3 (supporting). SpliceAI predicts no significant splicing alteration (max delta=0.04).

No variant-specific functional data, de novo observations, case-control enrichment, segregation data, or clinical phenotype information are available for this variant. Two publications (PMID:24345752 and PMID:34429321) were identified as potentially relevant but neither could be verified to contain this exact variant (NM_022552.4:c.2395C>T/p.Pro799Ser). PVS1 does not apply to this missense variant. No same-residue comparator variants (PM5) were identified. The variant has been observed once in COSMIC (COSV104391761) as a somatic finding.

With PM2 (supporting) and PP3 (supporting) as the only met criteria, the evidence does not reach any classification threshold under the generic ACMG/AMP 2015 combination rules (PMID:25741868). Two supporting criteria alone are insufficient for Likely Pathogenic (requires ≥1 moderate + 4 supporting or stronger combinations). This variant is classified as a Variant of Uncertain Significance (VUS). Review of full-text PMID:34429321 (Huang et al., systematic DNMT3A variant profiling) is recommended, as it may contain functional data for Pro799Ser that could upgrade PS3/BS3 status.