SCMCP

Instructions

Implementation Reference

• src/scmcp/tool/pp.py:120-137 (handler)

  Generic handler that executes the 'combat' tool by retrieving sc.pp.combat from pp_func, preparing arguments (including inplace=True), calling it on the active adata, handling errors, and logging the operation.

  def run_pp_func(ads, func, arguments):
      adata = ads.adata_dic[ads.active]
      if func not in pp_func:
          raise ValueError(f"不支持的函数: {func}")
      
      run_func = pp_func[func]
      parameters = inspect.signature(run_func).parameters
      arguments["inplace"] = True
      kwargs = {k: arguments.get(k) for k in parameters if k in arguments}
      try:
          res = run_func(adata, **kwargs)
          add_op_log(adata, run_func, kwargs)
      except KeyError as e:
          raise KeyError(f"Can not foud {e} column in adata.obs or adata.var")
      except Exception as e:
         raise e
      return res

• src/scmcp/schema/pp.py:363-390 (schema)

  Pydantic model defining the input schema for the 'combat' tool: 'key' (default 'batch') for the batch annotation column in adata.obs, and optional 'covariates' list of additional adjustment variables.

  class CombatModel(JSONParsingModel):
      """Input schema for the combat batch effect correction tool."""
      
      key: str = Field(
          default='batch',
          description="Key to a categorical annotation from adata.obs that will be used for batch effect removal."
      )
      
      covariates: Optional[List[str]] = Field(
          default=None,
          description="Additional covariates besides the batch variable such as adjustment variables or biological condition."
      )
      
      @field_validator('key')
      def validate_key(cls, v: str) -> str:
          """Validate key is not empty"""
          if not v.strip():
              raise ValueError("key cannot be empty")
          return v
      
      @field_validator('covariates')
      def validate_covariates(cls, v: Optional[List[str]]) -> Optional[List[str]]:
          """Validate covariates are non-empty strings if provided"""
          if v is not None:
              if not all(isinstance(item, str) and item.strip() for item in v):
                  raise ValueError("covariates must be non-empty strings")
          return v

• src/scmcp/tool/pp.py:69-73 (registration)

  Creates the MCP Tool instance for 'combat' with description and input schema from CombatModel.

  combat = types.Tool(
      name="combat",
      description="ComBat function for batch effect correction",
      inputSchema=CombatModel.model_json_schema(),
  )

• src/scmcp/tool/pp.py:88-101 (helper)

  Dictionary mapping tool names to their corresponding scanpy preprocessing functions; 'combat' maps to sc.pp.combat, the core batch correction function.

  pp_func = {
      "filter_genes": sc.pp.filter_genes,
      "filter_cells": sc.pp.filter_cells,
      "calculate_qc_metrics": partial(sc.pp.calculate_qc_metrics, inplace=True),
      "log1p": sc.pp.log1p,
      "normalize_total": sc.pp.normalize_total,
      "pca": sc.pp.pca,
      "highly_variable_genes": sc.pp.highly_variable_genes,
      "regress_out": sc.pp.regress_out,
      "scale": sc.pp.scale,
      "combat": sc.pp.combat,
      "scrublet": sc.pp.scrublet,
      "neighbors": sc.pp.neighbors,
  }

• src/scmcp/tool/pp.py:103-117 (registration)

  Registers all preprocessing tools including 'combat' in the pp_tools dictionary for tool dispatching.

  # 模型与函数名称的映射
  pp_tools = {
      "filter_genes": filter_genes,
      "filter_cells": filter_cells,
      "calculate_qc_metrics": calculate_qc_metrics,
      "log1p": log1p,
      "normalize_total": normalize_total,
      "pca": pca,
      "highly_variable_genes": highly_variable_genes,
      "regress_out": regress_out,
      "scale": scale,
      "combat": combat,
      "scrublet": scrublet,
      "neighbors": neighbors,
  }