You are a scientific literature expert evaluating the genetic association between {gene_name} and {phenotype}.

Below is a list of scientific articles about {gene_name} and/or {phenotype}:

{alias_note}{dismiss_note}

{articles_list}

Classify the gene–phenotype relationship into ONE of four categories:

1. **Established**  
   Strong, replicated evidence that variants in {gene_name} cause or are associated with {phenotype}. Multiple human studies in independent probands, families, or cohorts support a clear causal or associative role.
   Research used to prove this category has to be in Human.

2. **Existing**
   At least one credible study reports that variation in {gene_name} contributes to or is associated with {phenotype}, but evidence is limited or not definitive.  
   Also choose **Existing** when variants in {gene_name} are associated with a phenotype that is a well-recognized precursor, subtype, mechanism, or risk factor strongly linked to {phenotype}.
   Research used to prove this category has to be in Human.

3. **Hypothesized**  
   Literature suggests a possible relationship based on biology, even without direct evidence from genetic variants.  
   Includes expression, functional data, pathway links, or relevant animal models.  
   Excludes cases where gene and phenotype are merely co-mentioned without relevance.

4. **Novel**  
   No evidence that variants or biological activity of {gene_name} influence {phenotype}.  
   Co-mentions are incidental and provide no logical basis.

Your task:
1. Read the article titles and abstracts.  
2. Choose the best category.  
3. Provide a brief justification (3–5 sentences).  
4. Cite up to 10 relevant PMIDs from the provided list only.

Return your answer in **this exact XML format**:

<response>
  <verdict>Established|Existing|Hypothesized|Novel</verdict>
  <justification>Your brief explanation (3–5 sentences)</justification>
  <pmids>
    <pmid>12345678</pmid>
    ...
  </pmids>
</response>

Additional rules:
- If uncertain between two categories, choose the *less novel* one.  
- Hypothesized can apply without phenotype measurements if the gene’s biology is directly relevant.
- Think step by step internally: if the studies used to prove "Existing" or "Established" are not in human but in an animal or in vitro experiments, downgrade to "Hypothesized".
- Maximum 10 PMIDs.

IMPORTANT: before answering Established, name the articles that demonstrate the link in their own results, with variants found in people. If you cannot cite at least two such PMIDs, reporting independent probands, families, or cohorts, answer Existing at most. A copy number change spanning several genes, such as a deletion, duplication or microduplication, does not count towards those two, because it cannot single out this gene from its neighbours. An article that restates the association as known background, while its own results are about something else, is a citation and not evidence, and any number of such statements does not add up to replication.
